bookloupe: bookloupe/bookloupe.c@d22d8cd4f628 (annotated)

ali@0	1	/*************************************************************************/
ali@40	2	/* bookloupe--check for assorted weirdnesses in a PG candidate text file */
ali@68	3	/* */
ali@68	4	/* Copyright 2000-2005 Jim Tinsley <jtinsley@pobox.com> */
ali@68	5	/* Copyright 2012- J. Ali Harlow <ali@juiblex.co.uk> */
ali@68	6	/* */
ali@0	7	/* This program is free software; you can redistribute it and/or modify */
ali@0	8	/* it under the terms of the GNU General Public License as published by */
ali@0	9	/* the Free Software Foundation; either version 2 of the License, or */
ali@68	10	/* (at your option) any later version. */
ali@68	11	/* */
ali@0	12	/* This program is distributed in the hope that it will be useful, */
ali@68	13	/* but WITHOUT ANY WARRANTY; without even the implied warranty of */
ali@68	14	/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */
ali@68	15	/* GNU General Public License for more details. */
ali@68	16	/* */
ali@68	17	/* You should have received a copy of the GNU General Public License */
ali@68	18	/* along with this program. If not, see <http://www.gnu.org/licenses/>. */
ali@0	19	/*************************************************************************/
ali@0	20
ali@0	21	#include <stdio.h>
ali@0	22	#include <stdlib.h>
ali@0	23	#include <string.h>
ali@0	24	#include <ctype.h>
ali@73	25	#ifdef __WIN32__
ali@73	26	#include <windows.h>
ali@73	27	#endif
ali@69	28	#include <glib.h>
ali@69	29	#include <bl/bl.h>
ali@92	30	#include "bookloupe.h"
ali@92	31	#include "counters.h"
ali@93	32	#include "pending.h"
ali@71	33	#include "HTMLentities.h"
ali@0	34
ali@103	35	gchar charset; / Or NULL for auto (ISO_8859-1/ASCII or UNICODE) */
ali@103	36	GIConv charset_validator=(GIConv)-1;
ali@103	37
ali@69	38	gchar *prevline;
ali@0	39
ali@40	40	/* Common typos. */
ali@40	41	char *typo[] = {
ali@40	42	"teh", "th", "og", "fi", "ro", "adn", "yuo", "ot", "fo", "thet", "ane",
ali@40	43	"nad", "te", "ig", "acn", "ahve", "alot", "anbd", "andt", "awya", "aywa",
ali@40	44	"bakc", "om", "btu", "byt", "cna", "cxan", "coudl", "dont", "didnt",
ali@40	45	"couldnt", "wouldnt", "doesnt", "shouldnt", "doign", "ehr", "hmi", "hse",
ali@40	46	"esle", "eyt", "fitrs", "firts", "foudn", "frmo", "fromt", "fwe", "gaurd",
ali@40	47	"gerat", "goign", "gruop", "haev", "hda", "hearign", "seeign", "sayign",
ali@40	48	"herat", "hge", "hsa", "hsi", "hte", "htere", "htese", "htey", "htis",
ali@40	49	"hvae", "hwich", "idae", "ihs", "iits", "int", "iwll", "iwth", "jsut",
ali@40	50	"loev", "sefl", "myu", "nkow", "nver", "nwe", "nwo", "ocur", "ohter",
ali@40	51	"omre", "onyl", "otehr", "otu", "owrk", "owuld", "peice", "peices",
ali@40	52	"peolpe", "peopel", "perhasp", "perhpas", "pleasent", "poeple", "porblem",
ali@40	53	"porblems", "rwite", "saidt", "saidh", "saids", "seh", "smae", "smoe",
ali@40	54	"sohw", "stnad", "stopry", "stoyr", "stpo", "tahn", "taht", "tath",
ali@40	55	"tehy", "tghe", "tghis", "theri", "theyll", "thgat", "thge", "thier",
ali@40	56	"thna", "thne", "thnig", "thnigs", "thsi", "thsoe", "thta", "timne",
ali@40	57	"tirne", "tkae", "tthe", "tyhat", "tyhe", "veyr", "vou", "vour", "vrey",
ali@40	58	"waht", "wasnt", "awtn", "watn", "wehn", "whic", "whcih", "whihc", "whta",
ali@40	59	"wihch", "wief", "wiht", "witha", "wiull", "wnat", "wnated", "wnats",
ali@40	60	"woh", "wohle", "wokr", "woudl", "wriet", "wrod", "wroet", "wroking",
ali@40	61	"wtih", "wuould", "wya", "yera", "yeras", "yersa", "yoiu", "youve",
ali@40	62	"ytou", "yuor", "abead", "ahle", "ahout", "ahove", "altbough", "balf",
ali@40	63	"bardly", "bas", "bave", "baving", "bebind", "beld", "belp", "belped",
ali@40	64	"ber", "bere", "bim", "bis", "bome", "bouse", "bowever", "buge",
ali@40	65	"dehates", "deht", "han", "hecause", "hecome", "heen", "hefore", "hegan",
ali@40	66	"hegin", "heing", "helieve", "henefit", "hetter", "hetween", "heyond",
ali@40	67	"hig", "higber", "huild", "huy", "hy", "jobn", "joh", "meanwbile",
ali@40	68	"memher", "memhers", "numher", "numhers", "perbaps", "prohlem", "puhlic",
ali@40	69	"witbout", "arn", "hin", "hirn", "wrok", "wroked", "amd", "aud",
ali@40	70	"prornise", "prornised", "modem", "bo", "heside", "chapteb", "chaptee",
ali@40	71	"se", ""
ali@40	72	};
ali@0	73
ali@69	74	GTree *usertypo;
ali@0	75
ali@40	76	/* Common abbreviations and other OK words not to query as typos. */
ali@40	77	char *okword[] = {
ali@40	78	"mr", "mrs", "mss", "mssrs", "ft", "pm", "st", "dr", "hmm", "h'm", "hmmm",
ali@40	79	"rd", "sh", "br", "pp", "hm", "cf", "jr", "sr", "vs", "lb", "lbs", "ltd",
ali@40	80	"pompeii","hawaii","hawaiian", "hotbed", "heartbeat", "heartbeats",
ali@40	81	"outbid", "outbids", "frostbite", "frostbitten", ""
ali@40	82	};
ali@0	83
ali@40	84	/* Common abbreviations that cause otherwise unexplained periods. */
ali@40	85	char *abbrev[] = {
ali@40	86	"cent", "cents", "viz", "vol", "vols", "vid", "ed", "al", "etc", "op",
ali@40	87	"cit", "deg", "min", "chap", "oz", "mme", "mlle", "mssrs", ""
ali@40	88	};
ali@0	89
ali@40	90	/*
ali@40	91	* Two-Letter combinations that rarely if ever start words,
ali@40	92	* but are common scannos or otherwise common letter combinations.
ali@40	93	*/
ali@40	94	char *nostart[] = {
ali@40	95	"hr", "hl", "cb", "sb", "tb", "wb", "tl", "tn", "rn", "lt", "tj", ""
ali@40	96	};
ali@0	97
ali@40	98	/*
ali@40	99	* Two-Letter combinations that rarely if ever end words,
ali@40	100	* but are common scannos or otherwise common letter combinations.
ali@40	101	*/
ali@40	102	char *noend[] = {
ali@40	103	"cb", "gb", "pb", "sb", "tb", "wh", "fr", "br", "qu", "tw", "gl", "fl",
ali@40	104	"sw", "gr", "sl", "cl", "iy", ""
ali@40	105	};
ali@0	106
ali@40	107	char *markup[] = {
ali@40	108	"a", "b", "big", "blockquote", "body", "br", "center", "col", "div", "em",
ali@40	109	"font", "h1", "h2", "h3", "h4", "h5", "h6", "head", "hr", "html", "i",
ali@40	110	"img", "li", "meta", "ol", "p", "pre", "small", "span", "strong", "sub",
ali@40	111	"sup", "table", "td", "tfoot", "thead", "title", "tr", "tt", "u", "ul", ""
ali@40	112	};
ali@0	113
ali@40	114	char *DPmarkup[] = {
ali@40	115	"<sc>", "</sc>", "/", "/", "/#", "#/", "/$", "$/", "<tb>", ""
ali@40	116	};
ali@0	117
ali@40	118	char *nocomma[] = {
ali@40	119	"the", "it's", "their", "an", "mrs", "a", "our", "that's", "its", "whose",
ali@40	120	"every", "i'll", "your", "my", "mr", "mrs", "mss", "mssrs", "ft", "pm",
ali@40	121	"st", "dr", "rd", "pp", "cf", "jr", "sr", "vs", "lb", "lbs", "ltd", "i'm",
ali@40	122	"during", "let", "toward", "among", ""
ali@40	123	};
ali@0	124
ali@40	125	char *noperiod[] = {
ali@40	126	"every", "i'm", "during", "that's", "their", "your", "our", "my", "or",
ali@40	127	"and", "but", "as", "if", "the", "its", "it's", "until", "than", "whether",
ali@40	128	"i'll", "whose", "who", "because", "when", "let", "till", "very", "an",
ali@40	129	"among", "those", "into", "whom", "having", "thence", ""
ali@40	130	};
ali@0	131
ali@69	132	gboolean pswit[SWITNO]; /* program switches */
ali@103	133	gchar *opt_charset;
ali@0	134
ali@102	135	gboolean typo_compat,paranoid_compat;
ali@102	136
ali@69	137	static GOptionEntry options[]={
ali@69	138	{ "dp", 'd', 0, G_OPTION_ARG_NONE, pswit+DP_SWITCH,
ali@69	139	"Ignore DP-specific markup", NULL },
ali@102	140	{ "no-dp", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	141	G_OPTION_ARG_NONE, pswit+DP_SWITCH,
ali@102	142	"Don't ignore DP-specific markup", NULL },
ali@102	143	{ "echo", 0, G_OPTION_FLAG_HIDDEN, G_OPTION_ARG_NONE, pswit+ECHO_SWITCH,
ali@102	144	"Echo queried line", NULL },
ali@102	145	{ "no-echo", 'e', G_OPTION_FLAG_REVERSE,
ali@102	146	G_OPTION_ARG_NONE, pswit+ECHO_SWITCH,
ali@69	147	"Don't echo queried line", NULL },
ali@69	148	{ "squote", 's', 0, G_OPTION_ARG_NONE, pswit+SQUOTE_SWITCH,
ali@69	149	"Check single quotes", NULL },
ali@102	150	{ "no-squote", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	151	G_OPTION_ARG_NONE, pswit+SQUOTE_SWITCH,
ali@102	152	"Don't check single quotes", NULL },
ali@102	153	{ "typo", 0, 0, G_OPTION_ARG_NONE, pswit+TYPO_SWITCH,
ali@69	154	"Check common typos", NULL },
ali@102	155	{ "no-typo", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	156	G_OPTION_ARG_NONE, pswit+TYPO_SWITCH,
ali@102	157	"Don't check common typos", NULL },
ali@69	158	{ "qpara", 'p', 0, G_OPTION_ARG_NONE, pswit+QPARA_SWITCH,
ali@69	159	"Require closure of quotes on every paragraph", NULL },
ali@102	160	{ "no-qpara", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	161	G_OPTION_ARG_NONE, pswit+QPARA_SWITCH,
ali@102	162	"Don't require closure of quotes on every paragraph", NULL },
ali@102	163	{ "paranoid", 0, G_OPTION_FLAG_HIDDEN,
ali@102	164	G_OPTION_ARG_NONE, pswit+PARANOID_SWITCH,
ali@102	165	"Enable paranoid querying of everything", NULL },
ali@102	166	{ "no-paranoid", 0, G_OPTION_FLAG_REVERSE,
ali@102	167	G_OPTION_ARG_NONE, pswit+PARANOID_SWITCH,
ali@69	168	"Disable paranoid querying of everything", NULL },
ali@102	169	{ "line-end", 0, G_OPTION_FLAG_HIDDEN,
ali@102	170	G_OPTION_ARG_NONE, pswit+LINE_END_SWITCH,
ali@102	171	"Enable line end checking", NULL },
ali@102	172	{ "no-line-end", 'l', G_OPTION_FLAG_REVERSE,
ali@102	173	G_OPTION_ARG_NONE, pswit+LINE_END_SWITCH,
ali@69	174	"Disable line end checking", NULL },
ali@69	175	{ "overview", 'o', 0, G_OPTION_ARG_NONE, pswit+OVERVIEW_SWITCH,
ali@69	176	"Overview: just show counts", NULL },
ali@102	177	{ "no-overview", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	178	G_OPTION_ARG_NONE, pswit+OVERVIEW_SWITCH,
ali@102	179	"Show individual warnings", NULL },
ali@69	180	{ "stdout", 'y', 0, G_OPTION_ARG_NONE, pswit+STDOUT_SWITCH,
ali@69	181	"Output errors to stdout instead of stderr", NULL },
ali@102	182	{ "no-stdout", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	183	G_OPTION_ARG_NONE, pswit+STDOUT_SWITCH,
ali@102	184	"Output errors to stderr instead of stdout", NULL },
ali@69	185	{ "header", 'h', 0, G_OPTION_ARG_NONE, pswit+HEADER_SWITCH,
ali@69	186	"Echo header fields", NULL },
ali@102	187	{ "no-header", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	188	G_OPTION_ARG_NONE, pswit+HEADER_SWITCH,
ali@102	189	"Don't echo header fields", NULL },
ali@69	190	{ "markup", 'm', 0, G_OPTION_ARG_NONE, pswit+MARKUP_SWITCH,
ali@69	191	"Ignore markup in < >", NULL },
ali@102	192	{ "no-markup", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	193	G_OPTION_ARG_NONE, pswit+MARKUP_SWITCH,
ali@102	194	"No special handling for markup in < >", NULL },
ali@69	195	{ "usertypo", 'u', 0, G_OPTION_ARG_NONE, pswit+USERTYPO_SWITCH,
ali@69	196	"Use file of user-defined typos", NULL },
ali@102	197	{ "no-usertypo", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	198	G_OPTION_ARG_NONE, pswit+USERTYPO_SWITCH,
ali@102	199	"Ignore file of user-defined typos", NULL },
ali@102	200	{ "verbose", 'v', 0, G_OPTION_ARG_NONE, pswit+VERBOSE_SWITCH,
ali@102	201	"Verbose - list everything", NULL },
ali@102	202	{ "no-verbose", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@102	203	G_OPTION_ARG_NONE, pswit+VERBOSE_SWITCH,
ali@102	204	"Switch off verbose mode", NULL },
ali@103	205	{ "charset", 0, 0, G_OPTION_ARG_STRING, &opt_charset,
ali@103	206	"Set of characters valid for this ebook", "NAME" },
ali@102	207	{ NULL }
ali@102	208	};
ali@102	209
ali@102	210	/*
ali@102	211	* Options relating to configuration which make no sense from inside
ali@102	212	* a configuration file.
ali@102	213	*/
ali@102	214
ali@102	215	static GOptionEntry config_options[]={
ali@69	216	{ "web", 'w', 0, G_OPTION_ARG_NONE, pswit+WEB_SWITCH,
ali@69	217	"Defaults for use on www upload", NULL },
ali@102	218	{ "dump-config", 0, 0, G_OPTION_ARG_NONE, pswit+DUMP_CONFIG_SWITCH,
ali@102	219	"Dump current config settings", NULL },
ali@102	220	{ NULL }
ali@102	221	};
ali@102	222
ali@102	223	static GOptionEntry compatibility_options[]={
ali@102	224	{ "toggle-typo", 't', 0, G_OPTION_ARG_NONE, &typo_compat,
ali@102	225	"Toggle checking for common typos", NULL },
ali@102	226	{ "toggle-relaxed", 'x', 0, G_OPTION_ARG_NONE, &paranoid_compat,
ali@102	227	"Toggle both paranoid mode and common typos", NULL },
ali@69	228	{ NULL }
ali@69	229	};
ali@0	230
ali@94	231	long cnt_quote; /* for overview mode, count of quote queries */
ali@68	232	long cnt_brack; /* for overview mode, count of brackets queries */
ali@68	233	long cnt_bin; /* for overview mode, count of non-ASCII queries */
ali@68	234	long cnt_odd; /* for overview mode, count of odd character queries */
ali@68	235	long cnt_long; /* for overview mode, count of long line errors */
ali@68	236	long cnt_short; /* for overview mode, count of short line queries */
ali@68	237	long cnt_punct; /* for overview mode,
ali@68	238	count of punctuation and spacing queries */
ali@68	239	long cnt_dash; /* for overview mode, count of dash-related queries */
ali@68	240	long cnt_word; /* for overview mode, count of word queries */
ali@68	241	long cnt_html; /* for overview mode, count of html queries */
ali@68	242	long cnt_lineend; /* for overview mode, count of line-end queries */
ali@68	243	long cnt_spacend; /* count of lines with space at end */
ali@68	244	long linecnt; /* count of total lines in the file */
ali@68	245	long checked_linecnt; /* count of lines actually checked */
ali@0	246
ali@69	247	void proghelp(GOptionContext *context);
ali@69	248	void procfile(const char *);
ali@0	249
ali@69	250	gchar *running_from;
ali@0	251
ali@70	252	gboolean mixdigit(const char *);
ali@69	253	gchar getaword(const char *);
ali@101	254	char flgets(char *,long,int);
ali@0	255	void postprocess_for_HTML(char *);
ali@0	256	char linehasmarkup(char );
ali@0	257	char losemarkup(char );
ali@70	258	gboolean tagcomp(const char ,const char );
ali@71	259	void loseentities(char *);
ali@69	260	gboolean isroman(const char *);
ali@0	261	void postprocess_for_DP(char *);
ali@72	262	void print_as_windows_1252(const char *string);
ali@72	263	void print_as_utf_8(const char *string);
ali@0	264
ali@69	265	GTree qword,qperiod;
ali@68	266
ali@73	267	#ifdef __WIN32__
ali@73	268	UINT saved_cp;
ali@73	269	#endif
ali@73	270
ali@103	271	gboolean set_charset(const char name,GError *err)
ali@103	272	{
ali@103	273	/* The various UNICODE encodings all share the same character set. */
ali@103	274	const char *unicode_aliases[]={ "UCS-2", "UCS-2BE", "UCS-2LE", "UCS-4",
ali@103	275	"UCS-4BE", "UCS-4LE", "UCS2", "UCS4", "UNICODE", "UNICODEBIG",
ali@103	276	"UNICODELITTLE", "UTF-7", "UTF-8", "UTF-16", "UTF-16BE", "UTF-16LE",
ali@103	277	"UTF-32", "UTF-32BE", "UTF-32LE", "UTF7", "UTF8", "UTF16", "UTF16BE",
ali@103	278	"UTF16LE", "UTF32", "UTF32BE", "UTF32LE" };
ali@103	279	int i;
ali@103	280	if (charset)
ali@103	281	g_free(charset);
ali@103	282	if (charset_validator!=(GIConv)-1)
ali@103	283	g_iconv_close(charset_validator);
ali@103	284	if (!name \|\| !g_strcasecmp(name,"auto"))
ali@103	285	{
ali@103	286	charset=NULL;
ali@103	287	charset_validator=(GIConv)-1;
ali@103	288	return TRUE;
ali@103	289	}
ali@103	290	else
ali@103	291	charset=g_strdup(name);
ali@103	292	for(i=0;i<G_N_ELEMENTS(unicode_aliases);i++)
ali@103	293	if (!g_strcasecmp(charset,unicode_aliases[i]))
ali@103	294	{
ali@103	295	g_free(charset);
ali@103	296	charset=g_strdup("UTF-8");
ali@103	297	break;
ali@103	298	}
ali@103	299	if (!strcmp(charset,"UTF-8"))
ali@103	300	charset_validator=(GIConv)-1;
ali@103	301	else
ali@103	302	{
ali@103	303	charset_validator=g_iconv_open(charset,"UTF-8");
ali@103	304	if (charset_validator==(GIConv)-1)
ali@103	305	{
ali@103	306	g_set_error(err,G_CONVERT_ERROR,G_CONVERT_ERROR_NO_CONVERSION,
ali@103	307	"Unknown character set \"%s\"",charset);
ali@103	308	return FALSE;
ali@103	309	}
ali@103	310	}
ali@103	311	return TRUE;
ali@103	312	}
ali@103	313
ali@102	314	GKeyFile *config;
ali@102	315
ali@102	316	void config_file_update(GKeyFile *kf)
ali@102	317	{
ali@102	318	int i;
ali@103	319	const char *s;
ali@102	320	gboolean sw;
ali@102	321	for(i=0;options[i].long_name;i++)
ali@102	322	{
ali@102	323	if (g_str_has_prefix(options[i].long_name,"no-"))
ali@102	324	continue;
ali@102	325	if (options[i].arg==G_OPTION_ARG_NONE)
ali@102	326	{
ali@102	327	sw=(gboolean )options[i].arg_data;
ali@102	328	if (options[i].flags&G_OPTION_FLAG_REVERSE)
ali@102	329	sw=!sw;
ali@102	330	g_key_file_set_boolean(kf,"options",options[i].long_name,sw);
ali@102	331	}
ali@103	332	else if (options[i].arg==G_OPTION_ARG_STRING)
ali@103	333	{
ali@103	334	s=(gchar *)options[i].arg_data;
ali@103	335	if (!s)
ali@103	336	s="auto";
ali@103	337	g_key_file_set_string(kf,"options",options[i].long_name,s);
ali@103	338	}
ali@102	339	else
ali@102	340	g_assert_not_reached();
ali@102	341	}
ali@102	342	}
ali@102	343
ali@102	344	void config_file_add_comments(GKeyFile *kf)
ali@102	345	{
ali@102	346	int i;
ali@102	347	gchar *comment;
ali@102	348	g_key_file_set_comment(kf,NULL,NULL," Default configuration for bookloupe",
ali@102	349	NULL);
ali@102	350	for(i=0;options[i].long_name;i++)
ali@102	351	{
ali@102	352	if (g_str_has_prefix(options[i].long_name,"no-"))
ali@102	353	continue;
ali@102	354	comment=g_strconcat(" ",options[i].description,NULL);
ali@102	355	g_key_file_set_comment(kf,"options",options[i].long_name,comment,NULL);
ali@102	356	g_free(comment);
ali@102	357	}
ali@102	358	}
ali@102	359
ali@102	360	void dump_config(void)
ali@102	361	{
ali@102	362	gchar *s;
ali@102	363	if (config)
ali@102	364	config_file_update(config);
ali@102	365	else
ali@102	366	{
ali@102	367	config=g_key_file_new();
ali@102	368	config_file_update(config);
ali@102	369	config_file_add_comments(config);
ali@102	370	}
ali@102	371	s=g_key_file_to_data(config,NULL,NULL);
ali@102	372	if (s)
ali@102	373	g_print("%s",s);
ali@102	374	g_free(s);
ali@102	375	}
ali@102	376
ali@102	377	GKeyFile read_config_file(gchar *full_path)
ali@102	378	{
ali@102	379	int i;
ali@102	380	GError *err=NULL;
ali@102	381	gchar **search_dirs;
ali@102	382	gchar *path;
ali@102	383	const char *search_path;
ali@102	384	GKeyFile *kf;
ali@102	385	kf=g_key_file_new();
ali@102	386	search_path=g_getenv("BOOKLOUPE_CONFIG_PATH");
ali@102	387	if (search_path)
ali@102	388	{
ali@102	389	#ifdef __WIN32__
ali@102	390	search_dirs=g_strsplit(search_path,";",0);
ali@102	391	#else
ali@102	392	search_dirs=g_strsplit(search_path,":",0);
ali@102	393	#endif
ali@102	394	}
ali@102	395	else
ali@102	396	{
ali@102	397	search_dirs=g_new(gchar *,4);
ali@102	398	search_dirs[0]=g_get_current_dir();
ali@102	399	search_dirs[1]=g_strdup(running_from);
ali@102	400	search_dirs[2]=g_strdup(g_get_user_config_dir());
ali@102	401	search_dirs[3]=NULL;
ali@102	402	}
ali@102	403	for(i=0;search_dirs[i];i++)
ali@102	404	{
ali@102	405	path=g_build_filename(search_dirs[i],"bookloupe.ini",NULL);
ali@102	406	if (g_key_file_load_from_file(kf,path,
ali@102	407	G_KEY_FILE_KEEP_COMMENTS\|G_KEY_FILE_KEEP_TRANSLATIONS,&err))
ali@102	408	break;
ali@102	409	if (!g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@102	410	{
ali@102	411	g_printerr("Bookloupe: Error reading %s\n",path);
ali@102	412	g_printerr("%s\n",err->message);
ali@102	413	exit(1);
ali@102	414	}
ali@102	415	g_clear_error(&err);
ali@102	416	g_free(path);
ali@102	417	path=NULL;
ali@102	418	}
ali@102	419	if (!search_dirs[i])
ali@102	420	{
ali@102	421	g_key_file_free(kf);
ali@102	422	kf=NULL;
ali@102	423	}
ali@102	424	g_strfreev(search_dirs);
ali@102	425	if (full_path && kf)
ali@102	426	*full_path=path;
ali@102	427	else
ali@102	428	g_free(path);
ali@102	429	return kf;
ali@102	430	}
ali@102	431
ali@102	432	void parse_config_file(void)
ali@102	433	{
ali@102	434	int i,j;
ali@103	435	gchar path,s;
ali@102	436	gchar **keys;
ali@102	437	gboolean sw;
ali@102	438	GError *err=NULL;
ali@102	439	config=read_config_file(&path);
ali@102	440	if (config)
ali@102	441	keys=g_key_file_get_keys(config,"options",NULL,NULL);
ali@102	442	else
ali@102	443	keys=NULL;
ali@102	444	if (keys)
ali@102	445	{
ali@102	446	for(i=0;keys[i];i++)
ali@102	447	{
ali@102	448	for(j=0;options[j].long_name;j++)
ali@102	449	{
ali@102	450	if (g_str_has_prefix(options[j].long_name,"no-"))
ali@102	451	continue;
ali@102	452	else if (!strcmp(keys[i],options[j].long_name))
ali@102	453	{
ali@102	454	if (options[j].arg==G_OPTION_ARG_NONE)
ali@102	455	{
ali@102	456	sw=g_key_file_get_boolean(config,"options",keys[i],
ali@102	457	&err);
ali@102	458	if (err)
ali@102	459	{
ali@102	460	g_printerr("Bookloupe: %s: options.%s: %s\n",
ali@102	461	path,keys[i],err->message);
ali@102	462	g_clear_error(&err);
ali@102	463	}
ali@103	464	else
ali@103	465	{
ali@103	466	if (options[j].flags&G_OPTION_FLAG_REVERSE)
ali@103	467	sw=!sw;
ali@103	468	(gboolean )options[j].arg_data=sw;
ali@103	469	}
ali@103	470	break;
ali@103	471	}
ali@103	472	else if (options[j].arg==G_OPTION_ARG_STRING)
ali@103	473	{
ali@103	474	s=g_key_file_get_string(config,"options",keys[i],
ali@103	475	&err);
ali@103	476	if (err)
ali@103	477	{
ali@103	478	g_printerr("Bookloupe: %s: options.%s: %s\n",
ali@103	479	path,keys[i],err->message);
ali@103	480	g_clear_error(&err);
ali@103	481	}
ali@103	482	else
ali@103	483	{
ali@103	484	g_free((gchar *)options[j].arg_data);
ali@103	485	if (!g_strcmp0(s,"auto"))
ali@103	486	{
ali@103	487	(gchar *)options[j].arg_data=NULL;
ali@103	488	g_free(s);
ali@103	489	}
ali@103	490	else
ali@103	491	(gchar *)options[j].arg_data=s;
ali@103	492	}
ali@102	493	break;
ali@102	494	}
ali@102	495	else
ali@102	496	g_assert_not_reached();
ali@102	497	}
ali@102	498	}
ali@102	499	if (!options[j].long_name)
ali@102	500	g_printerr("Bookloupe: %s: Unknown option \"%s\" ignored\n",
ali@102	501	path,keys[i]);
ali@102	502	}
ali@102	503	g_strfreev(keys);
ali@102	504	}
ali@102	505	if (config)
ali@102	506	g_free(path);
ali@102	507	}
ali@102	508
ali@69	509	void parse_options(int argc,char **argv)
ali@0	510	{
ali@69	511	GError *err=NULL;
ali@69	512	GOptionContext *context;
ali@102	513	GOptionGroup *compatibility;
ali@69	514	context=g_option_context_new(
ali@102	515	"file - look for errors in Project Gutenberg(TM) etexts");
ali@69	516	g_option_context_add_main_entries(context,options,NULL);
ali@102	517	g_option_context_add_main_entries(context,config_options,NULL);
ali@102	518	compatibility=g_option_group_new("compatibility",
ali@102	519	"Options for Compatibility with Gutcheck:",
ali@102	520	"Show compatibility options",NULL,NULL);
ali@102	521	g_option_group_add_entries(compatibility,compatibility_options);
ali@102	522	g_option_context_add_group(context,compatibility);
ali@102	523	g_option_context_set_description(context,
ali@102	524	"For simplicity, only the switch options which reverse the\n"
ali@102	525	"default configuration are listed. In most cases, both vanilla\n"
ali@102	526	"and \"no-\" prefixed versions are available for use.");
ali@69	527	if (!g_option_context_parse(context,argc,argv,&err))
ali@69	528	{
ali@69	529	g_printerr("Bookloupe: %s\n",err->message);
ali@69	530	g_printerr("Use \"%s --help\" for help\n",(*argv)[0]);
ali@69	531	exit(1);
ali@69	532	}
ali@102	533	if (typo_compat)
ali@69	534	pswit[TYPO_SWITCH]=!pswit[TYPO_SWITCH];
ali@102	535	if (paranoid_compat)
ali@102	536	{
ali@102	537	pswit[PARANOID_SWITCH]=!pswit[PARANOID_SWITCH];
ali@102	538	pswit[TYPO_SWITCH]=!pswit[TYPO_SWITCH];
ali@102	539	}
ali@40	540	/*
ali@40	541	* Web uploads - for the moment, this is really just a placeholder
ali@40	542	* until we decide what processing we really want to do on web uploads
ali@40	543	*/
ali@40	544	if (pswit[WEB_SWITCH])
ali@40	545	{
ali@40	546	/* specific override for web uploads */
ali@69	547	pswit[ECHO_SWITCH]=TRUE;
ali@69	548	pswit[SQUOTE_SWITCH]=FALSE;
ali@69	549	pswit[TYPO_SWITCH]=TRUE;
ali@69	550	pswit[QPARA_SWITCH]=FALSE;
ali@69	551	pswit[PARANOID_SWITCH]=TRUE;
ali@69	552	pswit[LINE_END_SWITCH]=FALSE;
ali@69	553	pswit[OVERVIEW_SWITCH]=FALSE;
ali@69	554	pswit[STDOUT_SWITCH]=FALSE;
ali@69	555	pswit[HEADER_SWITCH]=TRUE;
ali@69	556	pswit[VERBOSE_SWITCH]=FALSE;
ali@69	557	pswit[MARKUP_SWITCH]=FALSE;
ali@69	558	pswit[USERTYPO_SWITCH]=FALSE;
ali@69	559	pswit[DP_SWITCH]=FALSE;
ali@40	560	}
ali@103	561	if (opt_charset && !set_charset(opt_charset,&err))
ali@103	562	{
ali@103	563	g_printerr("%s\n",err->message);
ali@103	564	exit(1);
ali@103	565	}
ali@102	566	if (pswit[DUMP_CONFIG_SWITCH])
ali@102	567	{
ali@102	568	dump_config();
ali@102	569	exit(0);
ali@102	570	}
ali@103	571	g_free(opt_charset);
ali@103	572	opt_charset=NULL;
ali@102	573	if (pswit[OVERVIEW_SWITCH])
ali@102	574	/* just print summary; don't echo */
ali@102	575	pswit[ECHO_SWITCH]=FALSE;
ali@69	576	if (*argc<2)
ali@40	577	{
ali@69	578	proghelp(context);
ali@69	579	exit(1);
ali@40	580	}
ali@69	581	g_option_context_free(context);
ali@69	582	}
ali@69	583
ali@69	584	/*
ali@69	585	* read_user_scannos:
ali@69	586	*
ali@69	587	* Read in the user-defined stealth scanno list.
ali@69	588	*/
ali@69	589	void read_user_scannos(void)
ali@69	590	{
ali@69	591	GError *err=NULL;
ali@69	592	gchar *usertypo_file;
ali@69	593	gboolean okay;
ali@69	594	int i;
ali@70	595	gsize len,nb;
ali@70	596	gchar contents,utf8,**lines;
ali@69	597	usertypo_file=g_strdup("bookloupe.typ");
ali@69	598	okay=file_get_contents_text(usertypo_file,&contents,&len,&err);
ali@69	599	if (g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@69	600	{
ali@69	601	g_clear_error(&err);
ali@69	602	g_free(usertypo_file);
ali@69	603	usertypo_file=g_build_filename(running_from,"bookloupe.typ",NULL);
ali@69	604	okay=file_get_contents_text(usertypo_file,&contents,&len,&err);
ali@69	605	}
ali@69	606	if (g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@69	607	{
ali@69	608	g_clear_error(&err);
ali@69	609	g_free(usertypo_file);
ali@69	610	usertypo_file=g_strdup("gutcheck.typ");
ali@69	611	okay=file_get_contents_text(usertypo_file,&contents,&len,&err);
ali@69	612	}
ali@69	613	if (g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@69	614	{
ali@69	615	g_clear_error(&err);
ali@69	616	g_free(usertypo_file);
ali@69	617	usertypo_file=g_build_filename(running_from,"gutcheck.typ",NULL);
ali@69	618	okay=file_get_contents_text(usertypo_file,&contents,&len,&err);
ali@69	619	}
ali@69	620	if (g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@69	621	{
ali@69	622	g_free(usertypo_file);
ali@70	623	g_print(" --> I couldn't find bookloupe.typ "
ali@69	624	"-- proceeding without user typos.\n");
ali@69	625	return;
ali@69	626	}
ali@69	627	else if (!okay)
ali@69	628	{
ali@69	629	fprintf(stderr,"%s: %s\n",usertypo_file,err->message);
ali@69	630	g_free(usertypo_file);
ali@69	631	g_clear_error(&err);
ali@69	632	exit(1);
ali@69	633	}
ali@72	634	if (g_utf8_validate(contents,len,NULL))
ali@103	635	{
ali@72	636	utf8=g_utf8_normalize(contents,len,G_NORMALIZE_DEFAULT_COMPOSE);
ali@103	637	if (!charset)
ali@103	638	(void)set_charset("UNICODE",NULL);
ali@103	639	}
ali@72	640	else
ali@72	641	utf8=g_convert(contents,len,"UTF-8","WINDOWS-1252",NULL,&nb,NULL);
ali@70	642	g_free(contents);
ali@70	643	lines=g_strsplit_set(utf8,"\r\n",0);
ali@70	644	g_free(utf8);
ali@69	645	usertypo=g_tree_new_full((GCompareDataFunc)strcmp,NULL,g_free,NULL);
ali@69	646	for (i=0;lines[i];i++)
ali@69	647	if ((unsigned char )lines[i]>'!')
ali@69	648	g_tree_insert(usertypo,lines[i],GINT_TO_POINTER(1));
ali@69	649	else
ali@69	650	g_free(lines[i]);
ali@69	651	g_free(lines);
ali@69	652	}
ali@69	653
ali@69	654	/*
ali@69	655	* read_etext:
ali@69	656	*
ali@69	657	* Read an etext returning a newly allocated string containing the file
ali@69	658	* contents or NULL on error.
ali@69	659	*/
ali@69	660	gchar read_etext(const char filename,GError **err)
ali@69	661	{
ali@76	662	GError *tmp_err=NULL;
ali@70	663	gchar contents,utf8;
ali@76	664	gsize len,bytes_read,bytes_written;
ali@76	665	int i,line,col;
ali@69	666	if (!g_file_get_contents(filename,&contents,&len,err))
ali@69	667	return NULL;
ali@72	668	if (g_utf8_validate(contents,len,NULL))
ali@72	669	{
ali@72	670	utf8=g_utf8_normalize(contents,len,G_NORMALIZE_DEFAULT_COMPOSE);
ali@72	671	g_set_print_handler(print_as_utf_8);
ali@73	672	#ifdef __WIN32__
ali@73	673	SetConsoleOutputCP(CP_UTF8);
ali@73	674	#endif
ali@72	675	}
ali@72	676	else
ali@72	677	{
ali@76	678	utf8=g_convert(contents,len,"UTF-8","WINDOWS-1252",&bytes_read,
ali@76	679	&bytes_written,&tmp_err);
ali@76	680	if (g_error_matches(tmp_err,G_CONVERT_ERROR,
ali@76	681	G_CONVERT_ERROR_ILLEGAL_SEQUENCE))
ali@76	682	{
ali@76	683	line=col=1;
ali@76	684	for(i=0;i<bytes_read;i++)
ali@76	685	if (contents[i]=='\n')
ali@76	686	{
ali@76	687	line++;
ali@76	688	col=1;
ali@76	689	}
ali@76	690	else if (contents[i]!='\r')
ali@76	691	col++;
ali@76	692	g_set_error(err,G_CONVERT_ERROR,G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
ali@76	693	"Input conversion failed. Byte %d at line %d, column %d is not a "
ali@76	694	"valid Windows-1252 character",
ali@76	695	((unsigned char *)contents)[bytes_read],line,col);
ali@76	696	}
ali@76	697	else if (tmp_err)
ali@76	698	g_propagate_error(err,tmp_err);
ali@72	699	g_set_print_handler(print_as_windows_1252);
ali@73	700	#ifdef __WIN32__
ali@73	701	SetConsoleOutputCP(1252);
ali@73	702	#endif
ali@72	703	}
ali@70	704	g_free(contents);
ali@70	705	return utf8;
ali@69	706	}
ali@69	707
ali@73	708	void cleanup_on_exit(void)
ali@73	709	{
ali@73	710	#ifdef __WIN32__
ali@73	711	SetConsoleOutputCP(saved_cp);
ali@73	712	#endif
ali@73	713	}
ali@73	714
ali@69	715	int main(int argc,char **argv)
ali@69	716	{
ali@73	717	#ifdef __WIN32__
ali@73	718	atexit(cleanup_on_exit);
ali@73	719	saved_cp=GetConsoleOutputCP();
ali@73	720	#endif
ali@69	721	running_from=g_path_get_dirname(argv[0]);
ali@102	722	/* Paranoid checking is turned OFF, not on, by its switch */
ali@102	723	pswit[PARANOID_SWITCH]=TRUE;
ali@102	724	/* if running in paranoid mode, typo checks default to enabled */
ali@102	725	pswit[TYPO_SWITCH]=TRUE;
ali@102	726	/* Line-end checking is turned OFF, not on, by its switch */
ali@102	727	pswit[LINE_END_SWITCH]=TRUE;
ali@102	728	/* Echoing is turned OFF, not on, by its switch */
ali@102	729	pswit[ECHO_SWITCH]=TRUE;
ali@102	730	parse_config_file();
ali@69	731	parse_options(&argc,&argv);
ali@40	732	if (pswit[USERTYPO_SWITCH])
ali@69	733	read_user_scannos();
ali@40	734	fprintf(stderr,"bookloupe: Check and report on an e-text\n");
ali@69	735	procfile(argv[1]);
ali@40	736	if (pswit[OVERVIEW_SWITCH])
ali@40	737	{
ali@70	738	g_print(" Checked %ld lines of %ld (head+foot = %ld)\n\n",
ali@40	739	checked_linecnt,linecnt,linecnt-checked_linecnt);
ali@70	740	g_print(" --------------- Queries found --------------\n");
ali@68	741	if (cnt_long)
ali@70	742	g_print(" Long lines: %14ld\n",cnt_long);
ali@68	743	if (cnt_short)
ali@70	744	g_print(" Short lines: %14ld\n",cnt_short);
ali@68	745	if (cnt_lineend)
ali@70	746	g_print(" Line-end problems: %14ld\n",cnt_lineend);
ali@68	747	if (cnt_word)
ali@70	748	g_print(" Common typos: %14ld\n",cnt_word);
ali@94	749	if (cnt_quote)
ali@94	750	g_print(" Unmatched quotes: %14ld\n",cnt_quote);
ali@68	751	if (cnt_brack)
ali@70	752	g_print(" Unmatched brackets: %14ld\n",cnt_brack);
ali@68	753	if (cnt_bin)
ali@70	754	g_print(" Non-ASCII characters: %14ld\n",cnt_bin);
ali@68	755	if (cnt_odd)
ali@70	756	g_print(" Proofing characters: %14ld\n",cnt_odd);
ali@68	757	if (cnt_punct)
ali@70	758	g_print(" Punctuation & spacing queries: %14ld\n",cnt_punct);
ali@68	759	if (cnt_dash)
ali@70	760	g_print(" Non-standard dashes: %14ld\n",cnt_dash);
ali@68	761	if (cnt_html)
ali@70	762	g_print(" Possible HTML tags: %14ld\n",cnt_html);
ali@70	763	g_print("\n");
ali@70	764	g_print(" TOTAL QUERIES %14ld\n",
ali@94	765	cnt_quote+cnt_brack+cnt_bin+cnt_odd+cnt_long+cnt_short+cnt_punct+
ali@94	766	cnt_dash+cnt_word+cnt_html+cnt_lineend);
ali@40	767	}
ali@69	768	g_free(running_from);
ali@69	769	if (usertypo)
ali@69	770	g_tree_unref(usertypo);
ali@103	771	set_charset(NULL,NULL);
ali@102	772	if (config)
ali@102	773	g_key_file_free(config);
ali@40	774	return 0;
ali@0	775	}
ali@0	776
ali@97	777	void count_dashes(const char line,const char dash,
ali@97	778	struct dash_results *results)
ali@97	779	{
ali@97	780	int i;
ali@97	781	gchar **tokens;
ali@97	782	gunichar pc,nc;
ali@97	783	gboolean spaced=FALSE,unspaced=FALSE,spaced2=FALSE;
ali@97	784	if (!*line)
ali@97	785	return;
ali@97	786	tokens=g_strsplit(line,dash,0);
ali@97	787	if (tokens[1])
ali@97	788	results->base++;
ali@97	789	for(i=1;tokens[i];i++)
ali@97	790	{
ali@97	791	pc=g_utf8_get_char(g_utf8_prev_char(tokens[i-1]+strlen(tokens[i-1])));
ali@97	792	nc=g_utf8_get_char(tokens[i]);
ali@97	793	if (g_unichar_isspace(pc) \|\| g_unichar_isspace(nc))
ali@97	794	spaced=TRUE;
ali@97	795	if (g_unichar_isspace(pc) && g_unichar_isspace(nc))
ali@97	796	spaced2=TRUE;
ali@97	797	else if (!g_unichar_isspace(pc) && !g_unichar_isspace(nc))
ali@97	798	unspaced=TRUE;
ali@97	799	}
ali@97	800	if (spaced)
ali@97	801	results->space++;
ali@97	802	if (spaced2)
ali@97	803	/* count of lines with em-dashes with spaces both sides */
ali@97	804	results->non_PG_space++;
ali@97	805	if (unspaced)
ali@97	806	/* count of lines with PG-type em-dashes with no spaces */
ali@97	807	results->PG_space++;
ali@97	808	g_strfreev(tokens);
ali@97	809	}
ali@97	810
ali@40	811	/*
ali@41	812	* first_pass:
ali@40	813	*
ali@41	814	* Run a first pass - verify that it's a valid PG
ali@41	815	* file, decide whether to report some things that
ali@41	816	* occur many times in the text like long or short
ali@41	817	* lines, non-standard dashes, etc.
ali@40	818	*/
ali@69	819	struct first_pass_results first_pass(const char etext)
ali@0	820	{
ali@70	821	gunichar laststart=CHAR_SPACE;
ali@54	822	const char *s;
ali@69	823	gchar *lc_line;
ali@70	824	int i,j,lbytes,llen;
ali@69	825	gchar **lines;
ali@41	826	unsigned int lastlen=0,lastblen=0;
ali@41	827	long spline=0,nspline=0;
ali@41	828	static struct first_pass_results results={0};
ali@97	829	struct dash_results tmp_dash_results;
ali@69	830	gchar *inword;
ali@94	831	QuoteClass qc;
ali@69	832	lines=g_strsplit(etext,"\n",0);
ali@101	833	if (!lines[0])
ali@101	834	{
ali@101	835	/* An empty etext has no terminators */
ali@101	836	results.newlines=DOS_NEWLINES;
ali@101	837	}
ali@101	838	else if (!lines[1])
ali@101	839	{
ali@101	840	/*
ali@101	841	* If there are no LFs, we don't have UNIX-style
ali@101	842	* terminators, but we might have OS9-style ones.
ali@101	843	*/
ali@101	844	results.newlines=OS9_NEWLINES;
ali@101	845	g_strfreev(lines);
ali@101	846	lines=g_strsplit(etext,"\r",0);
ali@101	847	if (!lines[0] \|\| !lines[1])
ali@101	848	/* Looks like we don't have any terminators at all */
ali@101	849	results.newlines=DOS_NEWLINES;
ali@101	850	}
ali@101	851	else
ali@101	852	{
ali@101	853	/* We might have UNIX-style terminators */
ali@101	854	results.newlines=UNIX_NEWLINES;
ali@101	855	}
ali@69	856	for (j=0;lines[j];j++)
ali@40	857	{
ali@70	858	lbytes=strlen(lines[j]);
ali@101	859	if (lbytes>0 && lines[j][lbytes-1]=='\r')
ali@101	860	{
ali@101	861	results.newlines=DOS_NEWLINES;
ali@101	862	do
ali@101	863	{
ali@101	864	lines[j][--lbytes]='\0';
ali@101	865	} while (lbytes>0 && lines[j][lbytes-1]=='\r');
ali@101	866	}
ali@70	867	llen=g_utf8_strlen(lines[j],lbytes);
ali@68	868	linecnt++;
ali@69	869	if (strstr(lines[j],"*END") && strstr(lines[j],"SMALL PRINT") &&
ali@69	870	(strstr(lines[j],"PUBLIC DOMAIN") \|\| strstr(lines[j],"COPYRIGHT")))
ali@40	871	{
ali@68	872	if (spline)
ali@70	873	g_print(" --> Duplicate header?\n");
ali@68	874	spline=linecnt+1; /* first line of non-header text, that is */
ali@40	875	}
ali@69	876	if (!strncmp(lines[j],"*** START",9) &&
ali@69	877	strstr(lines[j],"PROJECT GUTENBERG"))
ali@40	878	{
ali@68	879	if (nspline)
ali@70	880	g_print(" --> Duplicate header?\n");
ali@68	881	nspline=linecnt+1; /* first line of non-header text, that is */
ali@40	882	}
ali@68	883	if (spline \|\| nspline)
ali@40	884	{
ali@70	885	lc_line=g_utf8_strdown(lines[j],lbytes);
ali@69	886	if (strstr(lc_line,"end") && strstr(lc_line,"project gutenberg"))
ali@40	887	{
ali@69	888	if (strstr(lc_line,"end")<strstr(lc_line,"project gutenberg"))
ali@40	889	{
ali@68	890	if (results.footerline)
ali@40	891	{
ali@40	892	/* it's an old-form header - we can detect duplicates */
ali@68	893	if (!nspline)
ali@70	894	g_print(" --> Duplicate footer?\n");
ali@40	895	}
ali@68	896	else
ali@68	897	results.footerline=linecnt;
ali@40	898	}
ali@40	899	}
ali@69	900	g_free(lc_line);
ali@40	901	}
ali@68	902	if (spline)
ali@41	903	results.firstline=spline;
ali@68	904	if (nspline)
ali@41	905	results.firstline=nspline; /* override with new */
ali@68	906	if (results.footerline)
ali@40	907	continue; /* don't count the boilerplate in the footer */
ali@68	908	results.totlen+=llen;
ali@70	909	for (s=lines[j];*s;s=g_utf8_next_char(s))
ali@40	910	{
ali@70	911	if (g_utf8_get_char(s)>127)
ali@41	912	results.binlen++;
ali@70	913	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@41	914	results.alphalen++;
ali@94	915	if (s>lines[j])
ali@94	916	{
ali@94	917	if (CHAR_IS_DQUOTE(g_utf8_get_char(s)))
ali@94	918	qc=QUOTE_CLASS(g_utf8_get_char(s));
ali@94	919	else
ali@94	920	qc=INVALID_QUOTE;
ali@94	921	if ((qc==CLOSING_QUOTE \|\| qc==NEUTRAL_QUOTE) &&
ali@97	922	g_unichar_isalpha(g_utf8_get_char(g_utf8_prev_char(s))))
ali@94	923	results.endquote_count++;
ali@94	924	}
ali@40	925	}
ali@69	926	if (llen>2 && lastlen>2 && lastlen<SHORTEST_PG_LINE && lastblen>2 &&
ali@69	927	lastblen>SHORTEST_PG_LINE && laststart!=CHAR_SPACE)
ali@41	928	results.shortline++;
ali@70	929	if (lbytes>0 &&
ali@70	930	g_utf8_get_char(g_utf8_prev_char(lines[j]+lbytes))<=CHAR_SPACE)
ali@40	931	cnt_spacend++;
ali@69	932	if (strstr(lines[j],".,"))
ali@41	933	results.dotcomma++;
ali@68	934	/* only count ast lines for ignoring purposes where there is */
ali@68	935	/* locase text on the line */
ali@69	936	if (strchr(lines[j],'*'))
ali@40	937	{
ali@70	938	for (s=lines[j];*s;s=g_utf8_next_char(s))
ali@70	939	if (g_unichar_islower(g_utf8_get_char(s)))
ali@68	940	break;
ali@70	941	if (*s)
ali@41	942	results.astline++;
ali@40	943	}
ali@69	944	if (strchr(lines[j],'/'))
ali@68	945	results.fslashline++;
ali@82	946	if (lbytes>0)
ali@82	947	{
ali@82	948	for (s=g_utf8_prev_char(lines[j]+lbytes);
ali@82	949	s>lines[j] && g_utf8_get_char(s)<=CHAR_SPACE;
ali@82	950	s=g_utf8_prev_char(s))
ali@82	951	;
ali@82	952	if (s>g_utf8_next_char(lines[j]) && g_utf8_get_char(s)=='-' &&
ali@82	953	g_utf8_get_char(g_utf8_prev_char(s))!='-')
ali@82	954	results.hyphens++;
ali@82	955	}
ali@68	956	if (llen>LONGEST_PG_LINE)
ali@41	957	results.longline++;
ali@68	958	if (llen>WAY_TOO_LONG)
ali@41	959	results.verylongline++;
ali@69	960	if (strchr(lines[j],'<') && strchr(lines[j],'>'))
ali@40	961	{
ali@69	962	i=(int)(strchr(lines[j],'>')-strchr(lines[j],'<')+1);
ali@68	963	if (i>0)
ali@68	964	results.htmcount++;
ali@69	965	if (strstr(lines[j],"<i>"))
ali@41	966	results.htmcount+=4; /* bonus marks! */
ali@40	967	}
ali@68	968	/* Check for spaced em-dashes */
ali@97	969	memset(&tmp_dash_results,0,sizeof(tmp_dash_results));
ali@97	970	count_dashes(lines[j],"--",&tmp_dash_results);
ali@97	971	count_dashes(lines[j],"—",&tmp_dash_results);
ali@97	972	if (tmp_dash_results.base)
ali@97	973	results.emdash.base++;
ali@97	974	if (tmp_dash_results.non_PG_space)
ali@97	975	results.emdash.non_PG_space++;
ali@97	976	if (tmp_dash_results.PG_space)
ali@97	977	results.emdash.PG_space++;
ali@69	978	for (s=lines[j];*s;)
ali@40	979	{
ali@69	980	inword=getaword(&s);
ali@68	981	if (!strcmp(inword,"hij") \|\| !strcmp(inword,"niet"))
ali@68	982	results.Dutchcount++;
ali@68	983	if (!strcmp(inword,"dans") \|\| !strcmp(inword,"avec"))
ali@68	984	results.Frenchcount++;
ali@68	985	if (!strcmp(inword,"0") \|\| !strcmp(inword,"1"))
ali@68	986	results.standalone_digit++;
ali@69	987	g_free(inword);
ali@40	988	}
ali@68	989	/* Check for spaced dashes */
ali@69	990	if (strstr(lines[j]," -") && *(strstr(lines[j]," -")+2)!='-')
ali@41	991	results.spacedash++;
ali@68	992	lastblen=lastlen;
ali@69	993	lastlen=llen;
ali@69	994	laststart=lines[j][0];
ali@40	995	}
ali@69	996	g_strfreev(lines);
ali@41	997	return &results;
ali@41	998	}
ali@41	999
ali@42	1000	/*
ali@42	1001	* report_first_pass:
ali@42	1002	*
ali@42	1003	* Make some snap decisions based on the first pass results.
ali@42	1004	*/
ali@42	1005	struct warnings report_first_pass(struct first_pass_results results)
ali@42	1006	{
ali@42	1007	static struct warnings warnings={0};
ali@101	1008	warnings.newlines=results->newlines;
ali@101	1009	if (warnings.newlines==UNIX_NEWLINES)
ali@101	1010	g_print(" --> No lines in this file have a CR. Not reporting them. "
ali@101	1011	"Project Gutenberg requires that all lineends be CR-LF.\n");
ali@101	1012	else if (warnings.newlines==OS9_NEWLINES)
ali@101	1013	g_print(" --> No lines in this file have a LF. Not reporting them. "
ali@101	1014	"Project Gutenberg requires that all lineends be CR-LF.\n");
ali@42	1015	if (cnt_spacend>0)
ali@70	1016	g_print(" --> %ld lines in this file have white space at end\n",
ali@42	1017	cnt_spacend);
ali@42	1018	warnings.dotcomma=1;
ali@42	1019	if (results->dotcomma>5)
ali@42	1020	{
ali@68	1021	warnings.dotcomma=0;
ali@70	1022	g_print(" --> %ld lines in this file contain '.,'. "
ali@42	1023	"Not reporting them.\n",results->dotcomma);
ali@42	1024	}
ali@42	1025	/*
ali@42	1026	* If more than 50 lines, or one-tenth, are short,
ali@42	1027	* don't bother reporting them.
ali@42	1028	*/
ali@42	1029	warnings.shortline=1;
ali@42	1030	if (results->shortline>50 \|\| results->shortline*10>linecnt)
ali@42	1031	{
ali@68	1032	warnings.shortline=0;
ali@70	1033	g_print(" --> %ld lines in this file are short. "
ali@42	1034	"Not reporting short lines.\n",results->shortline);
ali@42	1035	}
ali@42	1036	/*
ali@42	1037	* If more than 50 lines, or one-tenth, are long,
ali@42	1038	* don't bother reporting them.
ali@42	1039	*/
ali@42	1040	warnings.longline=1;
ali@42	1041	if (results->longline>50 \|\| results->longline*10>linecnt)
ali@42	1042	{
ali@68	1043	warnings.longline=0;
ali@70	1044	g_print(" --> %ld lines in this file are long. "
ali@42	1045	"Not reporting long lines.\n",results->longline);
ali@42	1046	}
ali@42	1047	/* If more than 10 lines contain asterisks, don't bother reporting them. */
ali@42	1048	warnings.ast=1;
ali@42	1049	if (results->astline>10)
ali@42	1050	{
ali@68	1051	warnings.ast=0;
ali@70	1052	g_print(" --> %ld lines in this file contain asterisks. "
ali@42	1053	"Not reporting them.\n",results->astline);
ali@42	1054	}
ali@42	1055	/*
ali@42	1056	* If more than 10 lines contain forward slashes,
ali@42	1057	* don't bother reporting them.
ali@42	1058	*/
ali@42	1059	warnings.fslash=1;
ali@42	1060	if (results->fslashline>10)
ali@42	1061	{
ali@68	1062	warnings.fslash=0;
ali@70	1063	g_print(" --> %ld lines in this file contain forward slashes. "
ali@42	1064	"Not reporting them.\n",results->fslashline);
ali@42	1065	}
ali@42	1066	/*
ali@42	1067	* If more than 20 lines contain unpunctuated endquotes,
ali@42	1068	* don't bother reporting them.
ali@42	1069	*/
ali@42	1070	warnings.endquote=1;
ali@42	1071	if (results->endquote_count>20)
ali@42	1072	{
ali@68	1073	warnings.endquote=0;
ali@70	1074	g_print(" --> %ld lines in this file contain unpunctuated endquotes. "
ali@42	1075	"Not reporting them.\n",results->endquote_count);
ali@42	1076	}
ali@42	1077	/*
ali@42	1078	* If more than 15 lines contain standalone digits,
ali@42	1079	* don't bother reporting them.
ali@42	1080	*/
ali@42	1081	warnings.digit=1;
ali@42	1082	if (results->standalone_digit>10)
ali@42	1083	{
ali@68	1084	warnings.digit=0;
ali@70	1085	g_print(" --> %ld lines in this file contain standalone 0s and 1s. "
ali@42	1086	"Not reporting them.\n",results->standalone_digit);
ali@42	1087	}
ali@42	1088	/*
ali@42	1089	* If more than 20 lines contain hyphens at end,
ali@42	1090	* don't bother reporting them.
ali@42	1091	*/
ali@42	1092	warnings.hyphen=1;
ali@42	1093	if (results->hyphens>20)
ali@42	1094	{
ali@68	1095	warnings.hyphen=0;
ali@70	1096	g_print(" --> %ld lines in this file have hyphens at end. "
ali@42	1097	"Not reporting them.\n",results->hyphens);
ali@42	1098	}
ali@42	1099	if (results->htmcount>20 && !pswit[MARKUP_SWITCH])
ali@42	1100	{
ali@70	1101	g_print(" --> Looks like this is HTML. Switching HTML mode ON.\n");
ali@68	1102	pswit[MARKUP_SWITCH]=1;
ali@42	1103	}
ali@42	1104	if (results->verylongline>0)
ali@70	1105	g_print(" --> %ld lines in this file are VERY long!\n",
ali@42	1106	results->verylongline);
ali@42	1107	/*
ali@42	1108	* If there are more non-PG spaced dashes than PG em-dashes,
ali@42	1109	* assume it's deliberate.
ali@42	1110	* Current PG guidelines say don't use them, but older texts do,
ali@42	1111	* and some people insist on them whatever the guidelines say.
ali@42	1112	*/
ali@42	1113	warnings.dash=1;
ali@97	1114	if (results->spacedash+results->emdash.non_PG_space>
ali@97	1115	results->emdash.PG_space)
ali@42	1116	{
ali@68	1117	warnings.dash=0;
ali@70	1118	g_print(" --> There are %ld spaced dashes and em-dashes. "
ali@42	1119	"Not reporting them.\n",
ali@97	1120	results->spacedash+results->emdash.non_PG_space);
ali@42	1121	}
ali@103	1122	if (charset)
ali@103	1123	warnings.bin=0;
ali@103	1124	else
ali@42	1125	{
ali@103	1126	/* Charset ISO_8859-1/ASCII checks for compatibility with gutcheck */
ali@103	1127	warnings.bin=1;
ali@103	1128	/* If more than a quarter of characters are hi-bit, bug out. */
ali@103	1129	if (results->binlen*4>results->totlen)
ali@103	1130	{
ali@103	1131	g_print(" --> This file does not appear to be ASCII. "
ali@103	1132	"Terminating. Best of luck with it!\n");
ali@103	1133	exit(1);
ali@103	1134	}
ali@103	1135	if (results->alphalen*4<results->totlen)
ali@103	1136	{
ali@103	1137	g_print(" --> This file does not appear to be text. "
ali@103	1138	"Terminating. Best of luck with it!\n");
ali@103	1139	exit(1);
ali@103	1140	}
ali@103	1141	if (results->binlen*100>results->totlen \|\| results->binlen>100)
ali@103	1142	{
ali@103	1143	g_print(" --> There are a lot of foreign letters here. "
ali@103	1144	"Not reporting them.\n");
ali@103	1145	if (!pswit[VERBOSE_SWITCH])
ali@103	1146	warnings.bin=0;
ali@103	1147	}
ali@42	1148	}
ali@69	1149	warnings.isDutch=FALSE;
ali@42	1150	if (results->Dutchcount>50)
ali@42	1151	{
ali@69	1152	warnings.isDutch=TRUE;
ali@70	1153	g_print(" --> This looks like Dutch - "
ali@42	1154	"switching off dashes and warnings for 's Middags case.\n");
ali@42	1155	}
ali@69	1156	warnings.isFrench=FALSE;
ali@42	1157	if (results->Frenchcount>50)
ali@42	1158	{
ali@69	1159	warnings.isFrench=TRUE;
ali@70	1160	g_print(" --> This looks like French - "
ali@42	1161	"switching off some doublepunct.\n");
ali@42	1162	}
ali@42	1163	if (results->firstline && results->footerline)
ali@70	1164	g_print(" The PG header and footer appear to be already on.\n");
ali@42	1165	else
ali@42	1166	{
ali@68	1167	if (results->firstline)
ali@70	1168	g_print(" The PG header is on - no footer.\n");
ali@68	1169	if (results->footerline)
ali@70	1170	g_print(" The PG footer is on - no header.\n");
ali@42	1171	}
ali@70	1172	g_print("\n");
ali@42	1173	if (pswit[VERBOSE_SWITCH])
ali@42	1174	{
ali@68	1175	warnings.shortline=1;
ali@68	1176	warnings.dotcomma=1;
ali@68	1177	warnings.longline=1;
ali@68	1178	warnings.dash=1;
ali@68	1179	warnings.digit=1;
ali@68	1180	warnings.ast=1;
ali@68	1181	warnings.fslash=1;
ali@68	1182	warnings.hyphen=1;
ali@68	1183	warnings.endquote=1;
ali@70	1184	g_print(" * Verbose output is ON -- you asked for it! *\n");
ali@42	1185	}
ali@42	1186	if (warnings.isDutch)
ali@68	1187	warnings.dash=0;
ali@42	1188	if (results->footerline>0 && results->firstline>0 &&
ali@42	1189	results->footerline>results->firstline &&
ali@42	1190	results->footerline-results->firstline<100)
ali@42	1191	{
ali@70	1192	g_print(" --> I don't really know where this text starts. \n");
ali@70	1193	g_print(" There are no reference points.\n");
ali@70	1194	g_print(" I'm going to have to report the header and footer "
ali@42	1195	"as well.\n");
ali@68	1196	results->firstline=0;
ali@42	1197	}
ali@42	1198	return &warnings;
ali@42	1199	}
ali@42	1200
ali@43	1201	/*
ali@43	1202	* analyse_quotes:
ali@43	1203	*
ali@43	1204	* Look along the line, accumulate the count of quotes, and see
ali@43	1205	* if this is an empty line - i.e. a line with nothing on it
ali@43	1206	* but spaces.
ali@43	1207	* If line has just spaces, period, * and/or - on it, don't
ali@43	1208	* count it, since empty lines with asterisks or dashes to
ali@43	1209	* separate sections are common.
ali@43	1210	*
ali@69	1211	* Returns: TRUE if the line is empty.
ali@43	1212	*/
ali@98	1213	gboolean analyse_quotes(const char aline,struct counters counters)
ali@43	1214	{
ali@68	1215	int guessquote=0;
ali@69	1216	/* assume the line is empty until proven otherwise */
ali@69	1217	gboolean isemptyline=TRUE;
ali@70	1218	const char s=aline,sprev,*snext;
ali@70	1219	gunichar c;
ali@70	1220	sprev=NULL;
ali@94	1221	GError *tmp_err=NULL;
ali@43	1222	while (*s)
ali@43	1223	{
ali@70	1224	snext=g_utf8_next_char(s);
ali@70	1225	c=g_utf8_get_char(s);
ali@94	1226	if (CHAR_IS_DQUOTE(c))
ali@94	1227	(void)count_quote(counters,c,QUOTE_CLASS(c),&tmp_err);
ali@94	1228	else if (CHAR_IS_SQUOTE(c) && pswit[SQUOTE_SWITCH])
ali@43	1229	{
ali@43	1230	if (s==aline)
ali@43	1231	{
ali@43	1232	/*
ali@94	1233	* At start of line, it can only be a quotation mark.
ali@43	1234	* Hardcode a very common exception!
ali@43	1235	*/
ali@70	1236	if (!g_str_has_prefix(snext,"tis") &&
ali@70	1237	!g_str_has_prefix(snext,"Tis"))
ali@94	1238	(void)count_quote(counters,c,NEUTRAL_QUOTE,&tmp_err);
ali@43	1239	}
ali@70	1240	else if (g_unichar_isalpha(g_utf8_get_char(sprev)) &&
ali@70	1241	g_unichar_isalpha(g_utf8_get_char(snext)))
ali@43	1242	/* Do nothing! it's definitely an apostrophe, not a quote */
ali@43	1243	;
ali@43	1244	/* it's outside a word - let's check it out */
ali@92	1245	else if (c==CHAR_OPEN_SQUOTE \|\| c==CHAR_LS_QUOTE \|\|
ali@70	1246	g_unichar_isalpha(g_utf8_get_char(snext)))
ali@43	1247	{
ali@94	1248	/* certainly looks like a quotation mark */
ali@70	1249	if (!g_str_has_prefix(snext,"tis") &&
ali@70	1250	!g_str_has_prefix(snext,"Tis"))
ali@43	1251	/* hardcode a very common exception! */
ali@94	1252	{
ali@94	1253	if (strchr(".?!,;:",g_utf8_get_char(sprev)))
ali@94	1254	(void)count_quote(counters,c,NEUTRAL_QUOTE,&tmp_err);
ali@94	1255	else
ali@94	1256	(void)count_quote(counters,c,OPENING_QUOTE,&tmp_err);
ali@94	1257	}
ali@43	1258	}
ali@43	1259	else
ali@43	1260	{
ali@94	1261	/* now - is it a quotation mark? */
ali@43	1262	guessquote=0; /* accumulate clues */
ali@70	1263	if (g_unichar_isalpha(g_utf8_get_char(sprev)))
ali@43	1264	{
ali@43	1265	/* it follows a letter - could be either */
ali@43	1266	guessquote++;
ali@70	1267	if (g_utf8_get_char(sprev)=='s')
ali@43	1268	{
ali@43	1269	/* looks like a plural apostrophe */
ali@43	1270	guessquote-=3;
ali@70	1271	if (g_utf8_get_char(snext)==CHAR_SPACE)
ali@70	1272	/* bonus marks! */
ali@43	1273	guessquote-=2;
ali@43	1274	}
ali@94	1275	if (innermost_quote_matches(counters,c))
ali@94	1276	/*
ali@94	1277	* Give it the benefit of some doubt,
ali@94	1278	* if a squote is already open.
ali@94	1279	*/
ali@94	1280	guessquote++;
ali@94	1281	else
ali@94	1282	guessquote--;
ali@94	1283	if (guessquote>=0)
ali@94	1284	(void)count_quote(counters,c,CLOSING_QUOTE,&tmp_err);
ali@43	1285	}
ali@43	1286	else
ali@94	1287	/* no adjacent letter - it must be a quote of some kind */
ali@94	1288	(void)count_quote(counters,c,NEUTRAL_QUOTE,&tmp_err);
ali@43	1289	}
ali@43	1290	}
ali@94	1291	if (tmp_err)
ali@94	1292	{
ali@94	1293	if (pswit[ECHO_SWITCH])
ali@94	1294	g_print("\n%s\n",aline);
ali@94	1295	if (!pswit[OVERVIEW_SWITCH])
ali@94	1296	g_print(" Line %ld column %ld - %s\n",
ali@94	1297	linecnt,g_utf8_pointer_to_offset(aline,s)+1,tmp_err->message);
ali@94	1298	g_clear_error(&tmp_err);
ali@94	1299	}
ali@70	1300	if (c!=CHAR_SPACE && c!='-' && c!='.' && c!=CHAR_ASTERISK &&
ali@70	1301	c!='\r' && c!='\n')
ali@69	1302	isemptyline=FALSE; /* ignore lines like * * * as spacers */
ali@70	1303	if (c==CHAR_UNDERSCORE)
ali@43	1304	counters->c_unders++;
ali@93	1305	if (c==CHAR_OPEN_SBRACK)
ali@93	1306	{
ali@93	1307	if (!matching_difference(counters,COUNTER_ILLUSTRATION) &&
ali@93	1308	!matching_difference(counters,c) && s==aline &&
ali@93	1309	g_str_has_prefix(s,"[Illustration:"))
ali@93	1310	increment_matching(counters,COUNTER_ILLUSTRATION,TRUE);
ali@93	1311	else
ali@93	1312	increment_matching(counters,c,TRUE);
ali@93	1313	}
ali@93	1314	else if (c==CHAR_OPEN_CBRACK \|\| c==CHAR_OPEN_RBRACK)
ali@92	1315	increment_matching(counters,c,TRUE);
ali@93	1316	if (c==CHAR_CLOSE_SBRACK)
ali@93	1317	{
ali@93	1318	if (!matching_count(counters,COUNTER_ILLUSTRATION,FALSE) &&
ali@93	1319	!matching_difference(counters,c) && !*snext)
ali@93	1320	increment_matching(counters,COUNTER_ILLUSTRATION,FALSE);
ali@93	1321	else
ali@93	1322	increment_matching(counters,c,FALSE);
ali@93	1323	}
ali@93	1324	else if (c==CHAR_CLOSE_CBRACK \|\| c==CHAR_CLOSE_RBRACK)
ali@92	1325	increment_matching(counters,c,FALSE);
ali@70	1326	sprev=s;
ali@70	1327	s=snext;
ali@43	1328	}
ali@43	1329	return isemptyline;
ali@43	1330	}
ali@43	1331
ali@41	1332	/*
ali@67	1333	* check_for_control_characters:
ali@67	1334	*
ali@67	1335	* Check for invalid or questionable characters in the line
ali@67	1336	* Anything above 127 is invalid for plain ASCII, and
ali@67	1337	* non-printable control characters should also be flagged.
ali@67	1338	* Tabs should generally not be there.
ali@67	1339	*/
ali@67	1340	void check_for_control_characters(const char *aline)
ali@67	1341	{
ali@70	1342	gunichar c;
ali@67	1343	const char *s;
ali@70	1344	for (s=aline;*s;s=g_utf8_next_char(s))
ali@67	1345	{
ali@70	1346	c=g_utf8_get_char(s);
ali@67	1347	if (c<CHAR_SPACE && c!=CHAR_LF && c!=CHAR_CR && c!=CHAR_TAB)
ali@67	1348	{
ali@67	1349	if (pswit[ECHO_SWITCH])
ali@70	1350	g_print("\n%s\n",aline);
ali@67	1351	if (!pswit[OVERVIEW_SWITCH])
ali@70	1352	g_print(" Line %ld column %ld - Control character %u\n",
ali@70	1353	linecnt,g_utf8_pointer_to_offset(s,aline)+1,c);
ali@67	1354	else
ali@67	1355	cnt_bin++;
ali@67	1356	}
ali@67	1357	}
ali@67	1358	}
ali@67	1359
ali@67	1360	/*
ali@44	1361	* check_for_odd_characters:
ali@44	1362	*
ali@44	1363	* Check for binary and other odd characters.
ali@44	1364	*/
ali@44	1365	void check_for_odd_characters(const char aline,const struct warnings warnings,
ali@69	1366	gboolean isemptyline)
ali@44	1367	{
ali@44	1368	/* Don't repeat multiple warnings on one line. */
ali@103	1369	gboolean eInvalidChar=FALSE,eTab=FALSE,eTilde=FALSE;
ali@70	1370	gboolean eCarat=FALSE,eFSlash=FALSE,eAst=FALSE;
ali@44	1371	const char *s;
ali@70	1372	gunichar c;
ali@103	1373	gsize nb;
ali@103	1374	gchar *t;
ali@70	1375	for (s=aline;*s;s=g_utf8_next_char(s))
ali@44	1376	{
ali@70	1377	c=g_utf8_get_char(s);
ali@103	1378	if (warnings->bin && !eInvalidChar &&
ali@103	1379	(c<CHAR_SPACE && c!='\t' && c!='\n' \|\| c>127))
ali@44	1380	{
ali@44	1381	if (pswit[ECHO_SWITCH])
ali@70	1382	g_print("\n%s\n",aline);
ali@44	1383	if (!pswit[OVERVIEW_SWITCH])
ali@70	1384	if (c>127 && c<160 \|\| c>255)
ali@70	1385	g_print(" Line %ld column %ld - "
ali@70	1386	"Non-ISO-8859 character %u\n",
ali@70	1387	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@44	1388	else
ali@70	1389	g_print(" Line %ld column %ld - "
ali@70	1390	"Non-ASCII character %u\n",
ali@70	1391	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@44	1392	else
ali@44	1393	cnt_bin++;
ali@103	1394	eInvalidChar=TRUE;
ali@103	1395	}
ali@103	1396	if (!eInvalidChar && charset)
ali@103	1397	{
ali@103	1398	if (charset_validator==(GIConv)-1)
ali@103	1399	{
ali@103	1400	if (!g_unichar_isdefined(c))
ali@103	1401	{
ali@103	1402	if (pswit[ECHO_SWITCH])
ali@103	1403	g_print("\n%s\n",aline);
ali@103	1404	if (!pswit[OVERVIEW_SWITCH])
ali@103	1405	g_print(" Line %ld column %ld - Unassigned UNICODE "
ali@103	1406	"code point U+%04" G_GINT32_MODIFIER "X\n",
ali@103	1407	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@103	1408	else
ali@103	1409	cnt_bin++;
ali@103	1410	eInvalidChar=TRUE;
ali@103	1411	}
ali@103	1412	else if (c>=0xE000 && c<=0xF8FF \|\| c>=0xF0000 && c<=0xFFFFD \|\|
ali@103	1413	c>=100000 && c<=0x10FFFD)
ali@103	1414	{
ali@103	1415	if (pswit[ECHO_SWITCH])
ali@103	1416	g_print("\n%s\n",aline);
ali@103	1417	if (!pswit[OVERVIEW_SWITCH])
ali@103	1418	g_print(" Line %ld column %ld - Private Use "
ali@103	1419	"character U+%04" G_GINT32_MODIFIER "X\n",
ali@103	1420	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@103	1421	else
ali@103	1422	cnt_bin++;
ali@103	1423	eInvalidChar=TRUE;
ali@103	1424	}
ali@103	1425	}
ali@103	1426	else
ali@103	1427	{
ali@103	1428	t=g_convert_with_iconv(s,g_utf8_next_char(s)-s,
ali@103	1429	charset_validator,NULL,&nb,NULL);
ali@103	1430	if (t)
ali@103	1431	g_free(t);
ali@103	1432	else
ali@103	1433	{
ali@103	1434	if (pswit[ECHO_SWITCH])
ali@103	1435	g_print("\n%s\n",aline);
ali@103	1436	if (!pswit[OVERVIEW_SWITCH])
ali@103	1437	g_print(" Line %ld column %ld - Non-%s "
ali@103	1438	"character %u\n",linecnt,
ali@103	1439	g_utf8_pointer_to_offset(aline,s)+1,charset,c);
ali@103	1440	else
ali@103	1441	cnt_bin++;
ali@103	1442	eInvalidChar=TRUE;
ali@103	1443	}
ali@103	1444	}
ali@44	1445	}
ali@70	1446	if (!eTab && c==CHAR_TAB)
ali@44	1447	{
ali@44	1448	if (pswit[ECHO_SWITCH])
ali@70	1449	g_print("\n%s\n",aline);
ali@44	1450	if (!pswit[OVERVIEW_SWITCH])
ali@70	1451	g_print(" Line %ld column %ld - Tab character?\n",
ali@70	1452	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1453	else
ali@44	1454	cnt_odd++;
ali@70	1455	eTab=TRUE;
ali@44	1456	}
ali@70	1457	if (!eTilde && c==CHAR_TILDE)
ali@44	1458	{
ali@44	1459	/*
ali@44	1460	* Often used by OCR software to indicate an
ali@44	1461	* unrecognizable character.
ali@44	1462	*/
ali@44	1463	if (pswit[ECHO_SWITCH])
ali@70	1464	g_print("\n%s\n",aline);
ali@44	1465	if (!pswit[OVERVIEW_SWITCH])
ali@70	1466	g_print(" Line %ld column %ld - Tilde character?\n",
ali@70	1467	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1468	else
ali@44	1469	cnt_odd++;
ali@70	1470	eTilde=TRUE;
ali@44	1471	}
ali@70	1472	if (!eCarat && c==CHAR_CARAT)
ali@44	1473	{
ali@44	1474	if (pswit[ECHO_SWITCH])
ali@70	1475	g_print("\n%s\n",aline);
ali@44	1476	if (!pswit[OVERVIEW_SWITCH])
ali@70	1477	g_print(" Line %ld column %ld - Carat character?\n",
ali@70	1478	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1479	else
ali@44	1480	cnt_odd++;
ali@70	1481	eCarat=TRUE;
ali@44	1482	}
ali@70	1483	if (!eFSlash && c==CHAR_FORESLASH && warnings->fslash)
ali@44	1484	{
ali@44	1485	if (pswit[ECHO_SWITCH])
ali@70	1486	g_print("\n%s\n",aline);
ali@44	1487	if (!pswit[OVERVIEW_SWITCH])
ali@70	1488	g_print(" Line %ld column %ld - Forward slash?\n",
ali@70	1489	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1490	else
ali@44	1491	cnt_odd++;
ali@70	1492	eFSlash=TRUE;
ali@44	1493	}
ali@44	1494	/*
ali@44	1495	* Report asterisks only in paranoid mode,
ali@44	1496	* since they're often deliberate.
ali@44	1497	*/
ali@44	1498	if (!eAst && pswit[PARANOID_SWITCH] && warnings->ast && !isemptyline &&
ali@70	1499	c==CHAR_ASTERISK)
ali@44	1500	{
ali@44	1501	if (pswit[ECHO_SWITCH])
ali@70	1502	g_print("\n%s\n",aline);
ali@44	1503	if (!pswit[OVERVIEW_SWITCH])
ali@70	1504	g_print(" Line %ld column %ld - Asterisk?\n",
ali@70	1505	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1506	else
ali@44	1507	cnt_odd++;
ali@70	1508	eAst=TRUE;
ali@44	1509	}
ali@44	1510	}
ali@44	1511	}
ali@44	1512
ali@44	1513	/*
ali@45	1514	* check_for_long_line:
ali@45	1515	*
ali@45	1516	* Check for line too long.
ali@45	1517	*/
ali@45	1518	void check_for_long_line(const char *aline)
ali@45	1519	{
ali@70	1520	if (g_utf8_strlen(aline,-1)>LONGEST_PG_LINE)
ali@45	1521	{
ali@45	1522	if (pswit[ECHO_SWITCH])
ali@70	1523	g_print("\n%s\n",aline);
ali@45	1524	if (!pswit[OVERVIEW_SWITCH])
ali@70	1525	g_print(" Line %ld column %ld - Long line %ld\n",
ali@70	1526	linecnt,g_utf8_strlen(aline,-1),g_utf8_strlen(aline,-1));
ali@45	1527	else
ali@45	1528	cnt_long++;
ali@45	1529	}
ali@45	1530	}
ali@45	1531
ali@45	1532	/*
ali@45	1533	* check_for_short_line:
ali@45	1534	*
ali@45	1535	* Check for line too short.
ali@45	1536	*
ali@45	1537	* This one is a bit trickier to implement: we don't want to
ali@45	1538	* flag the last line of a paragraph for being short, so we
ali@45	1539	* have to wait until we know that our current line is a
ali@45	1540	* "normal" line, then report the _previous_ line if it was too
ali@45	1541	* short. We also don't want to report indented lines like
ali@45	1542	* chapter heads or formatted quotations. We therefore keep
ali@45	1543	* last->len as the length of the last line examined, and
ali@45	1544	* last->blen as the length of the last but one, and try to
ali@45	1545	* suppress unnecessary warnings by checking that both were of
ali@45	1546	* "normal" length. We keep the first character of the last
ali@45	1547	* line in last->start, and if it was a space, we assume that
ali@45	1548	* the formatting is deliberate. I can't figure out a way to
ali@45	1549	* distinguish something like a quoted verse left-aligned or
ali@45	1550	* the header or footer of a letter from a paragraph of short
ali@45	1551	* lines - maybe if I examined the whole paragraph, and if the
ali@45	1552	* para has less than, say, 8 lines and if all lines are short,
ali@45	1553	* then just assume it's OK? Need to look at some texts to see
ali@45	1554	* how often a formula like this would get the right result.
ali@45	1555	*/
ali@45	1556	void check_for_short_line(const char aline,const struct line_properties last)
ali@45	1557	{
ali@70	1558	if (g_utf8_strlen(aline,-1)>1 && last->len>1 &&
ali@70	1559	last->len<SHORTEST_PG_LINE && last->blen>1 &&
ali@70	1560	last->blen>SHORTEST_PG_LINE && last->start!=CHAR_SPACE)
ali@45	1561	{
ali@45	1562	if (pswit[ECHO_SWITCH])
ali@70	1563	g_print("\n%s\n",prevline);
ali@45	1564	if (!pswit[OVERVIEW_SWITCH])
ali@70	1565	g_print(" Line %ld column %ld - Short line %ld?\n",
ali@70	1566	linecnt-1,g_utf8_strlen(prevline,-1),g_utf8_strlen(prevline,-1));
ali@45	1567	else
ali@45	1568	cnt_short++;
ali@45	1569	}
ali@45	1570	}
ali@45	1571
ali@45	1572	/*
ali@46	1573	* check_for_starting_punctuation:
ali@46	1574	*
ali@46	1575	* Look for punctuation other than full ellipses at start of line.
ali@46	1576	*/
ali@46	1577	void check_for_starting_punctuation(const char *aline)
ali@46	1578	{
ali@70	1579	if (*aline && g_utf8_strchr(".?!,;:",-1,g_utf8_get_char(aline)) &&
ali@70	1580	!g_str_has_prefix(aline,". . ."))
ali@46	1581	{
ali@46	1582	if (pswit[ECHO_SWITCH])
ali@70	1583	g_print("\n%s\n",aline);
ali@46	1584	if (!pswit[OVERVIEW_SWITCH])
ali@70	1585	g_print(" Line %ld column 1 - Begins with punctuation?\n",
ali@46	1586	linecnt);
ali@46	1587	else
ali@46	1588	cnt_punct++;
ali@46	1589	}
ali@46	1590	}
ali@46	1591
ali@46	1592	/*
ali@97	1593	* str_emdash:
ali@97	1594	*
ali@97	1595	* Find the first em-dash, return a pointer to it and set <next> to the
ali@97	1596	* character following the dash.
ali@97	1597	*/
ali@97	1598	char str_emdash(const char s,const char **next)
ali@97	1599	{
ali@97	1600	const char s1,s2;
ali@97	1601	s1=strstr(s,"--");
ali@97	1602	s2=strstr(s,"—");
ali@97	1603	if (!s1)
ali@97	1604	{
ali@97	1605	if (s2)
ali@97	1606	*next=g_utf8_next_char(s2);
ali@97	1607	return (char *)s2;
ali@97	1608	}
ali@97	1609	else if (!s2)
ali@97	1610	{
ali@97	1611	*next=g_utf8_next_char(g_utf8_next_char(s1));
ali@97	1612	return (char *)s1;
ali@97	1613	}
ali@97	1614	else if (s1<s2)
ali@97	1615	{
ali@97	1616	*next=g_utf8_next_char(g_utf8_next_char(s1));
ali@97	1617	return (char *)s1;
ali@97	1618	}
ali@97	1619	else
ali@97	1620	{
ali@97	1621	*next=g_utf8_next_char(s2);
ali@97	1622	return (char *)s2;
ali@97	1623	}
ali@97	1624	}
ali@97	1625
ali@97	1626	/*
ali@47	1627	* check_for_spaced_emdash:
ali@47	1628	*
ali@47	1629	* Check for spaced em-dashes.
ali@47	1630	*
ali@97	1631	* We must check _all_ occurrences of em-dashes on the line
ali@97	1632	* hence the loop - even if the first dash is OK
ali@47	1633	* there may be another that's wrong later on.
ali@47	1634	*/
ali@47	1635	void check_for_spaced_emdash(const char *aline)
ali@47	1636	{
ali@70	1637	const char s,t,*next;
ali@97	1638	for (s=aline;t=str_emdash(s,&next);s=next)
ali@47	1639	{
ali@70	1640	if (t>aline && g_utf8_get_char(g_utf8_prev_char(t))==CHAR_SPACE \|\|
ali@70	1641	g_utf8_get_char(next)==CHAR_SPACE)
ali@47	1642	{
ali@47	1643	if (pswit[ECHO_SWITCH])
ali@70	1644	g_print("\n%s\n",aline);
ali@47	1645	if (!pswit[OVERVIEW_SWITCH])
ali@70	1646	g_print(" Line %ld column %ld - Spaced em-dash?\n",
ali@70	1647	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@47	1648	else
ali@47	1649	cnt_dash++;
ali@47	1650	}
ali@47	1651	}
ali@47	1652	}
ali@47	1653
ali@47	1654	/*
ali@47	1655	* check_for_spaced_dash:
ali@47	1656	*
ali@47	1657	* Check for spaced dashes.
ali@47	1658	*/
ali@47	1659	void check_for_spaced_dash(const char *aline)
ali@47	1660	{
ali@47	1661	const char *s;
ali@47	1662	if ((s=strstr(aline," -")))
ali@47	1663	{
ali@70	1664	if (g_utf8_get_char(g_utf8_next_char(g_utf8_next_char(s)))!='-')
ali@47	1665	{
ali@47	1666	if (pswit[ECHO_SWITCH])
ali@70	1667	g_print("\n%s\n",aline);
ali@47	1668	if (!pswit[OVERVIEW_SWITCH])
ali@70	1669	g_print(" Line %ld column %ld - Spaced dash?\n",
ali@70	1670	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@47	1671	else
ali@47	1672	cnt_dash++;
ali@47	1673	}
ali@47	1674	}
ali@47	1675	else if ((s=strstr(aline,"- ")))
ali@47	1676	{
ali@70	1677	if (s==aline \|\| g_utf8_get_char(g_utf8_prev_char(s))!='-')
ali@47	1678	{
ali@47	1679	if (pswit[ECHO_SWITCH])
ali@70	1680	g_print("\n%s\n",aline);
ali@47	1681	if (!pswit[OVERVIEW_SWITCH])
ali@70	1682	g_print(" Line %ld column %ld - Spaced dash?\n",
ali@70	1683	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@47	1684	else
ali@47	1685	cnt_dash++;
ali@47	1686	}
ali@47	1687	}
ali@47	1688	}
ali@47	1689
ali@47	1690	/*
ali@48	1691	* check_for_unmarked_paragraphs:
ali@48	1692	*
ali@48	1693	* Check for unmarked paragraphs indicated by separate speakers.
ali@48	1694	*
ali@48	1695	* May well be false positive:
ali@48	1696	* "Bravo!" "Wonderful!" called the crowd.
ali@48	1697	* but useful all the same.
ali@48	1698	*/
ali@48	1699	void check_for_unmarked_paragraphs(const char *aline)
ali@48	1700	{
ali@48	1701	const char *s;
ali@48	1702	s=strstr(aline,"\" \"");
ali@48	1703	if (!s)
ali@48	1704	s=strstr(aline,"\" \"");
ali@48	1705	if (s)
ali@48	1706	{
ali@48	1707	if (pswit[ECHO_SWITCH])
ali@70	1708	g_print("\n%s\n",aline);
ali@48	1709	if (!pswit[OVERVIEW_SWITCH])
ali@70	1710	g_print(" Line %ld column %ld - "
ali@70	1711	"Query missing paragraph break?\n",
ali@70	1712	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@48	1713	else
ali@48	1714	cnt_punct++;
ali@48	1715	}
ali@48	1716	}
ali@48	1717
ali@48	1718	/*
ali@49	1719	* check_for_jeebies:
ali@49	1720	*
ali@49	1721	* Check for "to he" and other easy h/b errors.
ali@49	1722	*
ali@49	1723	* This is a very inadequate effort on the h/b problem,
ali@49	1724	* but the phrase "to he" is always an error, whereas "to
ali@49	1725	* be" is quite common.
ali@49	1726	* Similarly, '"Quiet!", be said.' is a non-be error
ali@49	1727	* "to he" is _not_ always an error!:
ali@49	1728	* "Where they went to he couldn't say."
ali@49	1729	* Another false positive:
ali@49	1730	* What would "Cinderella" be without the . . .
ali@49	1731	* and another: "If he wants to he can see for himself."
ali@49	1732	*/
ali@49	1733	void check_for_jeebies(const char *aline)
ali@49	1734	{
ali@49	1735	const char *s;
ali@49	1736	s=strstr(aline," be could ");
ali@49	1737	if (!s)
ali@49	1738	s=strstr(aline," be would ");
ali@49	1739	if (!s)
ali@49	1740	s=strstr(aline," was be ");
ali@49	1741	if (!s)
ali@49	1742	s=strstr(aline," be is ");
ali@49	1743	if (!s)
ali@49	1744	s=strstr(aline," is be ");
ali@49	1745	if (!s)
ali@49	1746	s=strstr(aline,"\", be ");
ali@49	1747	if (!s)
ali@49	1748	s=strstr(aline,"\" be ");
ali@49	1749	if (!s)
ali@49	1750	s=strstr(aline,"\" be ");
ali@49	1751	if (!s)
ali@49	1752	s=strstr(aline," to he ");
ali@49	1753	if (s)
ali@49	1754	{
ali@49	1755	if (pswit[ECHO_SWITCH])
ali@70	1756	g_print("\n%s\n",aline);
ali@49	1757	if (!pswit[OVERVIEW_SWITCH])
ali@70	1758	g_print(" Line %ld column %ld - Query he/be error?\n",
ali@70	1759	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@49	1760	else
ali@49	1761	cnt_word++;
ali@49	1762	}
ali@49	1763	s=strstr(aline," the had ");
ali@49	1764	if (!s)
ali@49	1765	s=strstr(aline," a had ");
ali@49	1766	if (!s)
ali@49	1767	s=strstr(aline," they bad ");
ali@49	1768	if (!s)
ali@49	1769	s=strstr(aline," she bad ");
ali@49	1770	if (!s)
ali@49	1771	s=strstr(aline," he bad ");
ali@49	1772	if (!s)
ali@49	1773	s=strstr(aline," you bad ");
ali@49	1774	if (!s)
ali@49	1775	s=strstr(aline," i bad ");
ali@49	1776	if (s)
ali@49	1777	{
ali@49	1778	if (pswit[ECHO_SWITCH])
ali@70	1779	g_print("\n%s\n",aline);
ali@49	1780	if (!pswit[OVERVIEW_SWITCH])
ali@70	1781	g_print(" Line %ld column %ld - Query had/bad error?\n",
ali@70	1782	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@49	1783	else
ali@49	1784	cnt_word++;
ali@49	1785	}
ali@49	1786	s=strstr(aline,"; hut ");
ali@49	1787	if (!s)
ali@49	1788	s=strstr(aline,", hut ");
ali@49	1789	if (s)
ali@49	1790	{
ali@49	1791	if (pswit[ECHO_SWITCH])
ali@70	1792	g_print("\n%s\n",aline);
ali@49	1793	if (!pswit[OVERVIEW_SWITCH])
ali@70	1794	g_print(" Line %ld column %ld - Query hut/but error?\n",
ali@70	1795	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@49	1796	else
ali@49	1797	cnt_word++;
ali@49	1798	}
ali@49	1799	}
ali@49	1800
ali@49	1801	/*
ali@50	1802	* check_for_mta_from:
ali@50	1803	*
ali@50	1804	* Special case - angled bracket in front of "From" placed there by an
ali@50	1805	* MTA when sending an e-mail.
ali@50	1806	*/
ali@50	1807	void check_for_mta_from(const char *aline)
ali@50	1808	{
ali@50	1809	const char *s;
ali@50	1810	s=strstr(aline,">From");
ali@50	1811	if (s)
ali@50	1812	{
ali@50	1813	if (pswit[ECHO_SWITCH])
ali@70	1814	g_print("\n%s\n",aline);
ali@50	1815	if (!pswit[OVERVIEW_SWITCH])
ali@70	1816	g_print(" Line %ld column %ld - "
ali@70	1817	"Query angled bracket with From\n",
ali@70	1818	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@50	1819	else
ali@50	1820	cnt_punct++;
ali@50	1821	}
ali@50	1822	}
ali@50	1823
ali@50	1824	/*
ali@51	1825	* check_for_orphan_character:
ali@51	1826	*
ali@51	1827	* Check for a single character line -
ali@51	1828	* often an overflow from bad wrapping.
ali@51	1829	*/
ali@51	1830	void check_for_orphan_character(const char *aline)
ali@51	1831	{
ali@70	1832	gunichar c;
ali@70	1833	c=g_utf8_get_char(aline);
ali@70	1834	if (c && !*g_utf8_next_char(aline))
ali@51	1835	{
ali@70	1836	if (c=='I' \|\| c=='V' \|\| c=='X' \|\| c=='L' \|\| g_unichar_isdigit(c))
ali@51	1837	; /* Nothing - ignore numerals alone on a line. */
ali@51	1838	else
ali@51	1839	{
ali@51	1840	if (pswit[ECHO_SWITCH])
ali@70	1841	g_print("\n%s\n",aline);
ali@51	1842	if (!pswit[OVERVIEW_SWITCH])
ali@70	1843	g_print(" Line %ld column 1 - Query single character line\n",
ali@51	1844	linecnt);
ali@51	1845	else
ali@51	1846	cnt_punct++;
ali@51	1847	}
ali@51	1848	}
ali@51	1849	}
ali@51	1850
ali@51	1851	/*
ali@52	1852	* check_for_pling_scanno:
ali@52	1853	*
ali@52	1854	* Check for I" - often should be !
ali@52	1855	*/
ali@52	1856	void check_for_pling_scanno(const char *aline)
ali@52	1857	{
ali@52	1858	const char *s;
ali@52	1859	s=strstr(aline," I\"");
ali@52	1860	if (s)
ali@52	1861	{
ali@52	1862	if (pswit[ECHO_SWITCH])
ali@70	1863	g_print("\n%s\n",aline);
ali@52	1864	if (!pswit[OVERVIEW_SWITCH])
ali@70	1865	g_print(" Line %ld column %ld - Query I=exclamation mark?\n",
ali@70	1866	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@52	1867	else
ali@52	1868	cnt_punct++;
ali@52	1869	}
ali@52	1870	}
ali@52	1871
ali@52	1872	/*
ali@53	1873	* check_for_extra_period:
ali@53	1874	*
ali@53	1875	* Check for period without a capital letter. Cut-down from gutspell.
ali@53	1876	* Only works when it happens on a single line.
ali@53	1877	*/
ali@53	1878	void check_for_extra_period(const char aline,const struct warnings warnings)
ali@53	1879	{
ali@92	1880	const char s,t,s1,sprev;
ali@69	1881	int i;
ali@70	1882	gsize len;
ali@69	1883	gboolean istypo;
ali@69	1884	gchar *testword;
ali@92	1885	gunichar c,nc,pc,*decomposition;
ali@53	1886	if (pswit[PARANOID_SWITCH])
ali@53	1887	{
ali@70	1888	for (t=aline;t=strstr(t,". ");)
ali@53	1889	{
ali@69	1890	if (t==aline)
ali@53	1891	{
ali@70	1892	t=g_utf8_next_char(t);
ali@53	1893	/* start of line punctuation is handled elsewhere */
ali@53	1894	continue;
ali@53	1895	}
ali@70	1896	if (!g_unichar_isalpha(g_utf8_get_char(g_utf8_prev_char(t))))
ali@53	1897	{
ali@70	1898	t=g_utf8_next_char(t);
ali@53	1899	continue;
ali@53	1900	}
ali@53	1901	if (warnings->isDutch)
ali@53	1902	{
ali@53	1903	/* For Frank & Jeroen -- 's Middags case */
ali@70	1904	gunichar c2,c3,c4,c5;
ali@70	1905	c2=g_utf8_get_char(g_utf8_offset_to_pointer(t,2));
ali@70	1906	c3=g_utf8_get_char(g_utf8_offset_to_pointer(t,3));
ali@70	1907	c4=g_utf8_get_char(g_utf8_offset_to_pointer(t,4));
ali@70	1908	c5=g_utf8_get_char(g_utf8_offset_to_pointer(t,5));
ali@92	1909	if (CHAR_IS_APOSTROPHE(c2) &&
ali@92	1910	g_unichar_islower(c3) && c4==CHAR_SPACE &&
ali@92	1911	g_unichar_isupper(c5))
ali@53	1912	{
ali@70	1913	t=g_utf8_next_char(t);
ali@53	1914	continue;
ali@53	1915	}
ali@53	1916	}
ali@70	1917	s1=g_utf8_next_char(g_utf8_next_char(t));
ali@70	1918	while (*s1 && !g_unichar_isalpha(g_utf8_get_char(s1)) &&
ali@99	1919	!g_unichar_isdigit(g_utf8_get_char(s1)))
ali@70	1920	s1=g_utf8_next_char(s1);
ali@70	1921	if (g_unichar_islower(g_utf8_get_char(s1)))
ali@53	1922	{
ali@53	1923	/* we have something to investigate */
ali@69	1924	istypo=TRUE;
ali@53	1925	/* so let's go back and find out */
ali@92	1926	nc=g_utf8_get_char(t);
ali@92	1927	s1=g_utf8_prev_char(t);
ali@92	1928	c=g_utf8_get_char(s1);
ali@92	1929	sprev=g_utf8_prev_char(s1);
ali@92	1930	pc=g_utf8_get_char(sprev);
ali@92	1931	while (s1>=aline &&
ali@92	1932	(g_unichar_isalpha(c) \|\| g_unichar_isdigit(c) \|\|
ali@92	1933	g_unichar_isalpha(pc) && CHAR_IS_APOSTROPHE(c) &&
ali@92	1934	g_unichar_isalpha(nc)))
ali@92	1935	{
ali@92	1936	nc=c;
ali@92	1937	s1=sprev;
ali@92	1938	c=pc;
ali@92	1939	sprev=g_utf8_prev_char(s1);
ali@92	1940	pc=g_utf8_get_char(sprev);
ali@92	1941	}
ali@70	1942	s1=g_utf8_next_char(s1);
ali@69	1943	s=strchr(s1,'.');
ali@69	1944	if (s)
ali@69	1945	testword=g_strndup(s1,s-s1);
ali@69	1946	else
ali@69	1947	testword=g_strdup(s1);
ali@53	1948	for (i=0;*abbrev[i];i++)
ali@53	1949	if (!strcmp(testword,abbrev[i]))
ali@69	1950	istypo=FALSE;
ali@70	1951	if (g_unichar_isdigit(g_utf8_get_char(testword)))
ali@69	1952	istypo=FALSE;
ali@70	1953	if (!*g_utf8_next_char(testword))
ali@69	1954	istypo=FALSE;
ali@53	1955	if (isroman(testword))
ali@69	1956	istypo=FALSE;
ali@53	1957	if (istypo)
ali@53	1958	{
ali@69	1959	istypo=FALSE;
ali@70	1960	for (s=testword;*s;s=g_utf8_next_char(s))
ali@70	1961	{
ali@70	1962	decomposition=g_unicode_canonical_decomposition(
ali@70	1963	g_utf8_get_char(s),&len);
ali@70	1964	if (g_utf8_strchr("aeiou",-1,decomposition[0]))
ali@69	1965	istypo=TRUE;
ali@70	1966	g_free(decomposition);
ali@70	1967	}
ali@53	1968	}
ali@69	1969	if (istypo &&
ali@69	1970	(pswit[VERBOSE_SWITCH] \|\| !g_tree_lookup(qperiod,testword)))
ali@53	1971	{
ali@69	1972	g_tree_insert(qperiod,g_strdup(testword),
ali@69	1973	GINT_TO_POINTER(1));
ali@69	1974	if (pswit[ECHO_SWITCH])
ali@70	1975	g_print("\n%s\n",aline);
ali@69	1976	if (!pswit[OVERVIEW_SWITCH])
ali@70	1977	g_print(" Line %ld column %ld - Extra period?\n",
ali@70	1978	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@69	1979	else
ali@69	1980	cnt_punct++;
ali@53	1981	}
ali@69	1982	g_free(testword);
ali@53	1983	}
ali@70	1984	t=g_utf8_next_char(t);
ali@53	1985	}
ali@53	1986	}
ali@53	1987	}
ali@53	1988
ali@53	1989	/*
ali@54	1990	* check_for_following_punctuation:
ali@54	1991	*
ali@54	1992	* Check for words usually not followed by punctuation.
ali@54	1993	*/
ali@54	1994	void check_for_following_punctuation(const char *aline)
ali@54	1995	{
ali@54	1996	int i;
ali@54	1997	const char s,wordstart;
ali@70	1998	gunichar c;
ali@69	1999	gchar inword,t;
ali@54	2000	if (pswit[TYPO_SWITCH])
ali@54	2001	{
ali@54	2002	for (s=aline;*s;)
ali@54	2003	{
ali@54	2004	wordstart=s;
ali@69	2005	t=getaword(&s);
ali@69	2006	if (!*t)
ali@69	2007	{
ali@69	2008	g_free(t);
ali@54	2009	continue;
ali@69	2010	}
ali@70	2011	inword=g_utf8_strdown(t,-1);
ali@69	2012	g_free(t);
ali@54	2013	for (i=0;*nocomma[i];i++)
ali@54	2014	if (!strcmp(inword,nocomma[i]))
ali@54	2015	{
ali@70	2016	c=g_utf8_get_char(s);
ali@70	2017	if (c==',' \|\| c==';' \|\| c==':')
ali@54	2018	{
ali@54	2019	if (pswit[ECHO_SWITCH])
ali@70	2020	g_print("\n%s\n",aline);
ali@54	2021	if (!pswit[OVERVIEW_SWITCH])
ali@70	2022	g_print(" Line %ld column %ld - "
ali@54	2023	"Query punctuation after %s?\n",
ali@70	2024	linecnt,g_utf8_pointer_to_offset(aline,s)+1,
ali@70	2025	inword);
ali@54	2026	else
ali@54	2027	cnt_punct++;
ali@54	2028	}
ali@54	2029	}
ali@54	2030	for (i=0;*noperiod[i];i++)
ali@54	2031	if (!strcmp(inword,noperiod[i]))
ali@54	2032	{
ali@70	2033	c=g_utf8_get_char(s);
ali@70	2034	if (c=='.' \|\| c=='!')
ali@54	2035	{
ali@54	2036	if (pswit[ECHO_SWITCH])
ali@70	2037	g_print("\n%s\n",aline);
ali@54	2038	if (!pswit[OVERVIEW_SWITCH])
ali@70	2039	g_print(" Line %ld column %ld - "
ali@54	2040	"Query punctuation after %s?\n",
ali@70	2041	linecnt,g_utf8_pointer_to_offset(aline,s)+1,
ali@70	2042	inword);
ali@54	2043	else
ali@54	2044	cnt_punct++;
ali@54	2045	}
ali@54	2046	}
ali@69	2047	g_free(inword);
ali@54	2048	}
ali@54	2049	}
ali@54	2050	}
ali@54	2051
ali@54	2052	/*
ali@55	2053	* check_for_typos:
ali@55	2054	*
ali@55	2055	* Check for commonly mistyped words,
ali@55	2056	* and digits like 0 for O in a word.
ali@55	2057	*/
ali@55	2058	void check_for_typos(const char aline,struct warnings warnings)
ali@55	2059	{
ali@70	2060	const char s,t,nt,wordstart;
ali@70	2061	gchar *inword;
ali@70	2062	gunichar *decomposition;
ali@70	2063	gchar *testword;
ali@70	2064	int i,vowel,consonant,*dupcnt;
ali@70	2065	gboolean isdup,istypo,alower;
ali@92	2066	gunichar c,pc;
ali@70	2067	long offset,len;
ali@70	2068	gsize decomposition_len;
ali@55	2069	for (s=aline;*s;)
ali@55	2070	{
ali@55	2071	wordstart=s;
ali@69	2072	inword=getaword(&s);
ali@55	2073	if (!*inword)
ali@69	2074	{
ali@69	2075	g_free(inword);
ali@55	2076	continue; /* don't bother with empty lines */
ali@69	2077	}
ali@55	2078	if (mixdigit(inword))
ali@55	2079	{
ali@55	2080	if (pswit[ECHO_SWITCH])
ali@70	2081	g_print("\n%s\n",aline);
ali@55	2082	if (!pswit[OVERVIEW_SWITCH])
ali@70	2083	g_print(" Line %ld column %ld - Query digit in %s\n",
ali@70	2084	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+1,inword);
ali@55	2085	else
ali@55	2086	cnt_word++;
ali@55	2087	}
ali@55	2088	/*
ali@55	2089	* Put the word through a series of tests for likely typos and OCR
ali@55	2090	* errors.
ali@55	2091	*/
ali@69	2092	if (pswit[TYPO_SWITCH] \|\| pswit[USERTYPO_SWITCH])
ali@55	2093	{
ali@69	2094	istypo=FALSE;
ali@70	2095	alower=FALSE;
ali@70	2096	for (t=inword;*t;t=g_utf8_next_char(t))
ali@55	2097	{
ali@70	2098	c=g_utf8_get_char(t);
ali@70	2099	nt=g_utf8_next_char(t);
ali@55	2100	/* lowercase for testing */
ali@70	2101	if (g_unichar_islower(c))
ali@70	2102	alower=TRUE;
ali@70	2103	if (alower && (g_unichar_isupper(c) \|\| g_unichar_istitle(c)))
ali@55	2104	{
ali@55	2105	/*
ali@55	2106	* We have an uppercase mid-word. However, there are
ali@55	2107	* common cases:
ali@55	2108	* Mac and Mc like McGill
ali@55	2109	* French contractions like l'Abbe
ali@55	2110	*/
ali@70	2111	offset=g_utf8_pointer_to_offset(inword,t);
ali@92	2112	if (offset>0)
ali@92	2113	pc=g_utf8_get_char(g_utf8_prev_char(t));
ali@92	2114	else
ali@92	2115	pc='\0';
ali@70	2116	if (offset==2 && c=='m' && g_utf8_get_char(nt)=='c' \|\|
ali@70	2117	offset==3 && c=='m' && g_utf8_get_char(nt)=='a' &&
ali@70	2118	g_utf8_get_char(g_utf8_next_char(nt))=='c' \|\|
ali@92	2119	CHAR_IS_APOSTROPHE(pc))
ali@55	2120	; /* do nothing! */
ali@55	2121	else
ali@69	2122	istypo=TRUE;
ali@55	2123	}
ali@55	2124	}
ali@70	2125	testword=g_utf8_casefold(inword,-1);
ali@69	2126	}
ali@69	2127	if (pswit[TYPO_SWITCH])
ali@69	2128	{
ali@55	2129	/*
ali@55	2130	* Check for certain unlikely two-letter combinations at word
ali@55	2131	* start and end.
ali@55	2132	*/
ali@70	2133	len=g_utf8_strlen(testword,-1);
ali@70	2134	if (len>1)
ali@55	2135	{
ali@55	2136	for (i=0;*nostart[i];i++)
ali@70	2137	if (g_str_has_prefix(testword,nostart[i]))
ali@69	2138	istypo=TRUE;
ali@55	2139	for (i=0;*noend[i];i++)
ali@70	2140	if (g_str_has_suffix(testword,noend[i]))
ali@69	2141	istypo=TRUE;
ali@55	2142	}
ali@55	2143	/* ght is common, gbt never. Like that. */
ali@55	2144	if (strstr(testword,"cb"))
ali@69	2145	istypo=TRUE;
ali@55	2146	if (strstr(testword,"gbt"))
ali@69	2147	istypo=TRUE;
ali@55	2148	if (strstr(testword,"pbt"))
ali@69	2149	istypo=TRUE;
ali@55	2150	if (strstr(testword,"tbs"))
ali@69	2151	istypo=TRUE;
ali@55	2152	if (strstr(testword,"mrn"))
ali@69	2153	istypo=TRUE;
ali@55	2154	if (strstr(testword,"ahle"))
ali@69	2155	istypo=TRUE;
ali@55	2156	if (strstr(testword,"ihle"))
ali@69	2157	istypo=TRUE;
ali@55	2158	/*
ali@55	2159	* "TBE" does happen - like HEARTBEAT - but uncommon.
ali@55	2160	* Also "TBI" - frostbite, outbid - but uncommon.
ali@55	2161	* Similarly "ii" like Hawaii, or Pompeii, and in Roman
ali@55	2162	* numerals, but "ii" is a common scanno.
ali@55	2163	*/
ali@55	2164	if (strstr(testword,"tbi"))
ali@69	2165	istypo=TRUE;
ali@55	2166	if (strstr(testword,"tbe"))
ali@69	2167	istypo=TRUE;
ali@55	2168	if (strstr(testword,"ii"))
ali@69	2169	istypo=TRUE;
ali@55	2170	/*
ali@55	2171	* Check for no vowels or no consonants.
ali@55	2172	* If none, flag a typo.
ali@55	2173	*/
ali@70	2174	if (!istypo && len>1)
ali@55	2175	{
ali@55	2176	vowel=consonant=0;
ali@70	2177	for (t=testword;*t;t=g_utf8_next_char(t))
ali@55	2178	{
ali@70	2179	c=g_utf8_get_char(t);
ali@70	2180	decomposition=
ali@70	2181	g_unicode_canonical_decomposition(c,&decomposition_len);
ali@70	2182	if (c=='y' \|\| g_unichar_isdigit(c))
ali@55	2183	{
ali@55	2184	/* Yah, this is loose. */
ali@55	2185	vowel++;
ali@55	2186	consonant++;
ali@55	2187	}
ali@70	2188	else if (g_utf8_strchr("aeiou",-1,decomposition[0]))
ali@55	2189	vowel++;
ali@55	2190	else
ali@55	2191	consonant++;
ali@70	2192	g_free(decomposition);
ali@55	2193	}
ali@55	2194	if (!vowel \|\| !consonant)
ali@69	2195	istypo=TRUE;
ali@55	2196	}
ali@55	2197	/*
ali@55	2198	* Now exclude the word from being reported if it's in
ali@55	2199	* the okword list.
ali@55	2200	*/
ali@55	2201	for (i=0;*okword[i];i++)
ali@55	2202	if (!strcmp(testword,okword[i]))
ali@69	2203	istypo=FALSE;
ali@55	2204	/*
ali@55	2205	* What looks like a typo may be a Roman numeral.
ali@55	2206	* Exclude these.
ali@55	2207	*/
ali@55	2208	if (istypo && isroman(testword))
ali@69	2209	istypo=FALSE;
ali@55	2210	/* Check the manual list of typos. */
ali@55	2211	if (!istypo)
ali@55	2212	for (i=0;*typo[i];i++)
ali@55	2213	if (!strcmp(testword,typo[i]))
ali@69	2214	istypo=TRUE;
ali@55	2215	/*
ali@55	2216	* Check lowercase s, l, i and m - special cases.
ali@55	2217	* "j" - often a semi-colon gone wrong.
ali@55	2218	* "d" for a missing apostrophe - he d
ali@55	2219	* "n" for "in"
ali@55	2220	*/
ali@70	2221	if (!istypo && len==1 &&
ali@70	2222	g_utf8_strchr("slmijdn",-1,g_utf8_get_char(inword)))
ali@69	2223	istypo=TRUE;
ali@55	2224	if (istypo)
ali@55	2225	{
ali@69	2226	dupcnt=g_tree_lookup(qword,testword);
ali@69	2227	if (dupcnt)
ali@69	2228	{
ali@69	2229	(*dupcnt)++;
ali@69	2230	isdup=!pswit[VERBOSE_SWITCH];
ali@69	2231	}
ali@69	2232	else
ali@69	2233	{
ali@69	2234	dupcnt=g_new0(int,1);
ali@69	2235	g_tree_insert(qword,g_strdup(testword),dupcnt);
ali@69	2236	isdup=FALSE;
ali@69	2237	}
ali@55	2238	if (!isdup)
ali@55	2239	{
ali@55	2240	if (pswit[ECHO_SWITCH])
ali@70	2241	g_print("\n%s\n",aline);
ali@55	2242	if (!pswit[OVERVIEW_SWITCH])
ali@55	2243	{
ali@70	2244	g_print(" Line %ld column %ld - Query word %s",
ali@70	2245	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+1,
ali@70	2246	inword);
ali@69	2247	if (!pswit[VERBOSE_SWITCH])
ali@70	2248	g_print(" - not reporting duplicates");
ali@70	2249	g_print("\n");
ali@55	2250	}
ali@55	2251	else
ali@55	2252	cnt_word++;
ali@55	2253	}
ali@55	2254	}
ali@55	2255	}
ali@55	2256	/* check the user's list of typos */
ali@69	2257	if (!istypo && usertypo && g_tree_lookup(usertypo,testword))
ali@69	2258	{
ali@69	2259	if (pswit[ECHO_SWITCH])
ali@70	2260	g_print("\n%s\n",aline);
ali@69	2261	if (!pswit[OVERVIEW_SWITCH])
ali@70	2262	g_print(" Line %ld column %ld - Query possible scanno %s\n",
ali@70	2263	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+2,inword);
ali@69	2264	}
ali@69	2265	if (pswit[TYPO_SWITCH] \|\| pswit[USERTYPO_SWITCH])
ali@69	2266	g_free(testword);
ali@55	2267	if (pswit[PARANOID_SWITCH] && warnings->digit)
ali@55	2268	{
ali@55	2269	/* In paranoid mode, query all 0 and 1 standing alone. */
ali@55	2270	if (!strcmp(inword,"0") \|\| !strcmp(inword,"1"))
ali@55	2271	{
ali@55	2272	if (pswit[ECHO_SWITCH])
ali@70	2273	g_print("\n%s\n",aline);
ali@55	2274	if (!pswit[OVERVIEW_SWITCH])
ali@70	2275	g_print(" Line %ld column %ld - Query standalone %s\n",
ali@70	2276	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+2,
ali@70	2277	inword);
ali@55	2278	else
ali@55	2279	cnt_word++;
ali@55	2280	}
ali@55	2281	}
ali@69	2282	g_free(inword);
ali@55	2283	}
ali@55	2284	}
ali@55	2285
ali@56	2286	/*
ali@56	2287	* check_for_misspaced_punctuation:
ali@56	2288	*
ali@56	2289	* Look for added or missing spaces around punctuation and quotes.
ali@56	2290	* If there is a punctuation character like ! with no space on
ali@56	2291	* either side, suspect a missing!space. If there are spaces on
ali@56	2292	* both sides , assume a typo. If we see a double quote with no
ali@56	2293	* space or punctuation on either side of it, assume unspaced
ali@56	2294	* quotes "like"this.
ali@56	2295	*/
ali@56	2296	void check_for_misspaced_punctuation(const char *aline,
ali@69	2297	struct parities *parities,gboolean isemptyline)
ali@56	2298	{
ali@69	2299	gboolean isacro,isellipsis;
ali@56	2300	const char *s;
ali@70	2301	gunichar c,nc,pc,n2c;
ali@94	2302	int parity;
ali@70	2303	c=g_utf8_get_char(aline);
ali@70	2304	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2305	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2306	{
ali@70	2307	pc=c;
ali@70	2308	c=nc;
ali@70	2309	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2310	/* For each character in the line after the first. */
ali@70	2311	if (g_utf8_strchr(".?!,;:_",-1,c)) /* if it's punctuation */
ali@56	2312	{
ali@56	2313	/* we need to suppress warnings for acronyms like M.D. */
ali@69	2314	isacro=FALSE;
ali@56	2315	/* we need to suppress warnings for ellipsis . . . */
ali@69	2316	isellipsis=FALSE;
ali@70	2317	/*
ali@70	2318	* If there are letters on both sides of it or
ali@70	2319	* if it's strict punctuation followed by an alpha.
ali@70	2320	*/
ali@70	2321	if (g_unichar_isalpha(nc) && (g_unichar_isalpha(pc) \|\|
ali@70	2322	g_utf8_strchr("?!,;:",-1,c)))
ali@56	2323	{
ali@70	2324	if (c=='.')
ali@56	2325	{
ali@70	2326	if (g_utf8_pointer_to_offset(aline,s)>2 &&
ali@70	2327	g_utf8_get_char(g_utf8_offset_to_pointer(s,-2))=='.')
ali@69	2328	isacro=TRUE;
ali@70	2329	n2c=g_utf8_get_char(g_utf8_next_char(g_utf8_next_char(s)));
ali@70	2330	if (nc && n2c=='.')
ali@69	2331	isacro=TRUE;
ali@56	2332	}
ali@56	2333	if (!isacro)
ali@56	2334	{
ali@56	2335	if (pswit[ECHO_SWITCH])
ali@70	2336	g_print("\n%s\n",aline);
ali@56	2337	if (!pswit[OVERVIEW_SWITCH])
ali@70	2338	g_print(" Line %ld column %ld - Missing space?\n",
ali@70	2339	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2340	else
ali@56	2341	cnt_punct++;
ali@56	2342	}
ali@56	2343	}
ali@70	2344	if (pc==CHAR_SPACE && (nc==CHAR_SPACE \|\| !nc))
ali@56	2345	{
ali@56	2346	/*
ali@56	2347	* If there are spaces on both sides,
ali@56	2348	* or space before and end of line.
ali@56	2349	*/
ali@70	2350	if (c=='.')
ali@56	2351	{
ali@70	2352	if (g_utf8_pointer_to_offset(aline,s)>2 &&
ali@70	2353	g_utf8_get_char(g_utf8_offset_to_pointer(s,-2))=='.')
ali@69	2354	isellipsis=TRUE;
ali@70	2355	n2c=g_utf8_get_char(g_utf8_next_char(g_utf8_next_char(s)));
ali@70	2356	if (nc && n2c=='.')
ali@69	2357	isellipsis=TRUE;
ali@56	2358	}
ali@56	2359	if (!isemptyline && !isellipsis)
ali@56	2360	{
ali@56	2361	if (pswit[ECHO_SWITCH])
ali@70	2362	g_print("\n%s\n",aline);
ali@56	2363	if (!pswit[OVERVIEW_SWITCH])
ali@70	2364	g_print(" Line %ld column %ld - "
ali@70	2365	"Spaced punctuation?\n",linecnt,
ali@70	2366	g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2367	else
ali@56	2368	cnt_punct++;
ali@56	2369	}
ali@56	2370	}
ali@56	2371	}
ali@56	2372	}
ali@56	2373	/* Split out the characters that CANNOT be preceded by space. */
ali@70	2374	c=g_utf8_get_char(aline);
ali@70	2375	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2376	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2377	{
ali@70	2378	pc=c;
ali@70	2379	c=nc;
ali@70	2380	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2381	/* for each character in the line after the first */
ali@70	2382	if (g_utf8_strchr("?!,;:",-1,c))
ali@56	2383	{
ali@56	2384	/* if it's punctuation that _cannot_ have a space before it */
ali@70	2385	if (pc==CHAR_SPACE && !isemptyline && nc!=CHAR_SPACE)
ali@56	2386	{
ali@56	2387	/*
ali@70	2388	* If nc DOES == space,
ali@56	2389	* it was already reported just above.
ali@56	2390	*/
ali@56	2391	if (pswit[ECHO_SWITCH])
ali@70	2392	g_print("\n%s\n",aline);
ali@56	2393	if (!pswit[OVERVIEW_SWITCH])
ali@70	2394	g_print(" Line %ld column %ld - Spaced punctuation?\n",
ali@70	2395	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2396	else
ali@56	2397	cnt_punct++;
ali@56	2398	}
ali@56	2399	}
ali@56	2400	}
ali@56	2401	/*
ali@56	2402	* Special case " .X" where X is any alpha.
ali@56	2403	* This plugs a hole in the acronym code above.
ali@56	2404	* Inelegant, but maintainable.
ali@56	2405	*/
ali@70	2406	c=g_utf8_get_char(aline);
ali@70	2407	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2408	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2409	{
ali@70	2410	pc=c;
ali@70	2411	c=nc;
ali@70	2412	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2413	/* for each character in the line after the first */
ali@70	2414	if (c=='.')
ali@56	2415	{
ali@56	2416	/* if it's a period */
ali@70	2417	if (pc==CHAR_SPACE && g_unichar_isalpha(nc))
ali@56	2418	{
ali@56	2419	/*
ali@56	2420	* If the period follows a space and
ali@56	2421	* is followed by a letter.
ali@56	2422	*/
ali@56	2423	if (pswit[ECHO_SWITCH])
ali@70	2424	g_print("\n%s\n",aline);
ali@56	2425	if (!pswit[OVERVIEW_SWITCH])
ali@70	2426	g_print(" Line %ld column %ld - Spaced punctuation?\n",
ali@70	2427	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2428	else
ali@56	2429	cnt_punct++;
ali@56	2430	}
ali@56	2431	}
ali@56	2432	}
ali@70	2433	c=g_utf8_get_char(aline);
ali@70	2434	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2435	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2436	{
ali@70	2437	pc=c;
ali@70	2438	c=nc;
ali@70	2439	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2440	/* for each character in the line after the first */
ali@94	2441	if (CHAR_IS_DQUOTE(c))
ali@56	2442	{
ali@70	2443	if (!g_utf8_strchr(" _-.'`,;:!/([{?}])",-1,pc) &&
ali@70	2444	!g_utf8_strchr(" _-.'`,;:!/([{?}])",-1,nc) && nc \|\|
ali@70	2445	!g_utf8_strchr(" _-([{'`",-1,pc) && g_unichar_isalpha(nc))
ali@56	2446	{
ali@56	2447	if (pswit[ECHO_SWITCH])
ali@70	2448	g_print("\n%s\n",aline);
ali@56	2449	if (!pswit[OVERVIEW_SWITCH])
ali@70	2450	g_print(" Line %ld column %ld - Unspaced quotes?\n",
ali@70	2451	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2452	else
ali@56	2453	cnt_punct++;
ali@56	2454	}
ali@56	2455	}
ali@56	2456	}
ali@56	2457	/* Check parity of quotes. */
ali@70	2458	nc=g_utf8_get_char(aline);
ali@70	2459	for (s=aline;*s;s=g_utf8_next_char(s))
ali@56	2460	{
ali@70	2461	c=nc;
ali@70	2462	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@94	2463	if (CHAR_IS_DQUOTE(c))
ali@56	2464	{
ali@94	2465	if (c==CHAR_DQUOTE)
ali@94	2466	{
ali@94	2467	parities->dquote=!parities->dquote;
ali@94	2468	parity=parities->dquote;
ali@94	2469	}
ali@94	2470	else if (c==CHAR_LD_QUOTE)
ali@94	2471	parity=1;
ali@94	2472	else
ali@94	2473	parity=0;
ali@94	2474	if (!parity)
ali@56	2475	{
ali@56	2476	/* parity even */
ali@99	2477	if (!g_utf8_strchr("_-.'`‘’/,;:!?)]} ",-1,nc))
ali@56	2478	{
ali@56	2479	if (pswit[ECHO_SWITCH])
ali@70	2480	g_print("\n%s\n",aline);
ali@56	2481	if (!pswit[OVERVIEW_SWITCH])
ali@70	2482	g_print(" Line %ld column %ld - "
ali@70	2483	"Wrongspaced quotes?\n",
ali@70	2484	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2485	else
ali@56	2486	cnt_punct++;
ali@56	2487	}
ali@56	2488	}
ali@56	2489	else
ali@56	2490	{
ali@56	2491	/* parity odd */
ali@99	2492	if (!g_unichar_isalpha(nc) && !g_unichar_isdigit(nc) &&
ali@99	2493	!g_utf8_strchr("_-/.'`‘’([{$",-1,nc) \|\| !nc)
ali@56	2494	{
ali@56	2495	if (pswit[ECHO_SWITCH])
ali@70	2496	g_print("\n%s\n",aline);
ali@56	2497	if (!pswit[OVERVIEW_SWITCH])
ali@70	2498	g_print(" Line %ld column %ld - "
ali@70	2499	"Wrongspaced quotes?\n",
ali@70	2500	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2501	else
ali@56	2502	cnt_punct++;
ali@56	2503	}
ali@56	2504	}
ali@56	2505	}
ali@56	2506	}
ali@94	2507	c=g_utf8_get_char(aline);
ali@94	2508	if (CHAR_IS_DQUOTE(c))
ali@56	2509	{
ali@70	2510	if (g_utf8_strchr(",;:!?)]} ",-1,
ali@70	2511	g_utf8_get_char(g_utf8_next_char(aline))))
ali@56	2512	{
ali@56	2513	if (pswit[ECHO_SWITCH])
ali@70	2514	g_print("\n%s\n",aline);
ali@56	2515	if (!pswit[OVERVIEW_SWITCH])
ali@70	2516	g_print(" Line %ld column 1 - Wrongspaced quotes?\n",
ali@56	2517	linecnt);
ali@56	2518	else
ali@56	2519	cnt_punct++;
ali@56	2520	}
ali@56	2521	}
ali@56	2522	if (pswit[SQUOTE_SWITCH])
ali@56	2523	{
ali@70	2524	nc=g_utf8_get_char(aline);
ali@70	2525	for (s=aline;*s;s=g_utf8_next_char(s))
ali@56	2526	{
ali@70	2527	c=nc;
ali@70	2528	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@92	2529	if (CHAR_IS_SQUOTE(c) && (s==aline \|\| s>aline &&
ali@70	2530	!g_unichar_isalpha(g_utf8_get_char(g_utf8_prev_char(s))) \|\|
ali@70	2531	!g_unichar_isalpha(nc)))
ali@56	2532	{
ali@56	2533	parities->squote=!parities->squote;
ali@56	2534	if (!parities->squote)
ali@56	2535	{
ali@56	2536	/* parity even */
ali@70	2537	if (!g_utf8_strchr("_-.'`/\",;:!?)]} ",-1,nc))
ali@56	2538	{
ali@56	2539	if (pswit[ECHO_SWITCH])
ali@70	2540	g_print("\n%s\n",aline);
ali@56	2541	if (!pswit[OVERVIEW_SWITCH])
ali@70	2542	g_print(" Line %ld column %ld - "
ali@56	2543	"Wrongspaced singlequotes?\n",
ali@70	2544	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2545	else
ali@56	2546	cnt_punct++;
ali@56	2547	}
ali@56	2548	}
ali@56	2549	else
ali@56	2550	{
ali@56	2551	/* parity odd */
ali@99	2552	if (!g_unichar_isalpha(nc) && !g_unichar_isdigit(nc) &&
ali@70	2553	!g_utf8_strchr("_-/\".'`",-1,nc) \|\| !nc)
ali@56	2554	{
ali@56	2555	if (pswit[ECHO_SWITCH])
ali@70	2556	g_print("\n%s\n",aline);
ali@56	2557	if (!pswit[OVERVIEW_SWITCH])
ali@70	2558	g_print(" Line %ld column %ld - "
ali@56	2559	"Wrongspaced singlequotes?\n",
ali@70	2560	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2561	else
ali@56	2562	cnt_punct++;
ali@56	2563	}
ali@56	2564	}
ali@56	2565	}
ali@56	2566	}
ali@56	2567	}
ali@56	2568	}
ali@56	2569
ali@55	2570	/*
ali@57	2571	* check_for_double_punctuation:
ali@57	2572	*
ali@57	2573	* Look for double punctuation like ,. or ,,
ali@57	2574	* Thanks to DW for the suggestion!
ali@57	2575	* In books with references, ".," and ".;" are common
ali@57	2576	* e.g. "etc., etc.," and vol. 1.; vol 3.;
ali@57	2577	* OTOH, from my initial tests, there are also fairly
ali@57	2578	* common errors. What to do? Make these cases paranoid?
ali@57	2579	* ".," is the most common, so warnings->dotcomma is used
ali@57	2580	* to suppress detailed reporting if it occurs often.
ali@57	2581	*/
ali@57	2582	void check_for_double_punctuation(const char aline,struct warnings warnings)
ali@57	2583	{
ali@70	2584	const char *s;
ali@70	2585	gunichar c,nc;
ali@70	2586	nc=g_utf8_get_char(aline);
ali@70	2587	for (s=aline;*s;s=g_utf8_next_char(s))
ali@57	2588	{
ali@70	2589	c=nc;
ali@70	2590	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@57	2591	/* for each punctuation character in the line */
ali@70	2592	if (c && nc && g_utf8_strchr(".?!,;:",-1,c) &&
ali@70	2593	g_utf8_strchr(".?!,;:",-1,nc))
ali@57	2594	{
ali@57	2595	/* followed by punctuation, it's a query, unless . . . */
ali@70	2596	if (c==nc && (c=='.' \|\| c=='?' \|\| c=='!') \|\|
ali@70	2597	!warnings->dotcomma && c=='.' && nc==',' \|\|
ali@70	2598	warnings->isFrench && g_str_has_prefix(s,",...") \|\|
ali@70	2599	warnings->isFrench && g_str_has_prefix(s,"...,") \|\|
ali@70	2600	warnings->isFrench && g_str_has_prefix(s,";...") \|\|
ali@70	2601	warnings->isFrench && g_str_has_prefix(s,"...;") \|\|
ali@70	2602	warnings->isFrench && g_str_has_prefix(s,":...") \|\|
ali@70	2603	warnings->isFrench && g_str_has_prefix(s,"...:") \|\|
ali@70	2604	warnings->isFrench && g_str_has_prefix(s,"!...") \|\|
ali@70	2605	warnings->isFrench && g_str_has_prefix(s,"...!") \|\|
ali@70	2606	warnings->isFrench && g_str_has_prefix(s,"?...") \|\|
ali@70	2607	warnings->isFrench && g_str_has_prefix(s,"...?"))
ali@57	2608	{
ali@70	2609	if (warnings->isFrench && g_str_has_prefix(s,",...") \|\|
ali@70	2610	warnings->isFrench && g_str_has_prefix(s,"...,") \|\|
ali@70	2611	warnings->isFrench && g_str_has_prefix(s,";...") \|\|
ali@70	2612	warnings->isFrench && g_str_has_prefix(s,"...;") \|\|
ali@70	2613	warnings->isFrench && g_str_has_prefix(s,":...") \|\|
ali@70	2614	warnings->isFrench && g_str_has_prefix(s,"...:") \|\|
ali@70	2615	warnings->isFrench && g_str_has_prefix(s,"!...") \|\|
ali@70	2616	warnings->isFrench && g_str_has_prefix(s,"...!") \|\|
ali@70	2617	warnings->isFrench && g_str_has_prefix(s,"?...") \|\|
ali@70	2618	warnings->isFrench && g_str_has_prefix(s,"...?"))
ali@70	2619	{
ali@70	2620	s+=4;
ali@70	2621	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@70	2622	}
ali@57	2623	; /* do nothing for .. !! and ?? which can be legit */
ali@57	2624	}
ali@57	2625	else
ali@57	2626	{
ali@57	2627	if (pswit[ECHO_SWITCH])
ali@70	2628	g_print("\n%s\n",aline);
ali@57	2629	if (!pswit[OVERVIEW_SWITCH])
ali@70	2630	g_print(" Line %ld column %ld - Double punctuation?\n",
ali@70	2631	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@57	2632	else
ali@57	2633	cnt_punct++;
ali@57	2634	}
ali@57	2635	}
ali@57	2636	}
ali@57	2637	}
ali@57	2638
ali@57	2639	/*
ali@58	2640	* check_for_spaced_quotes:
ali@58	2641	*/
ali@58	2642	void check_for_spaced_quotes(const char *aline)
ali@58	2643	{
ali@92	2644	int i;
ali@58	2645	const char s,t;
ali@92	2646	const gunichar single_quotes[]={CHAR_SQUOTE,CHAR_OPEN_SQUOTE,CHAR_LS_QUOTE,
ali@92	2647	CHAR_RS_QUOTE};
ali@92	2648	GString *pattern;
ali@58	2649	s=aline;
ali@58	2650	while ((t=strstr(s," \" ")))
ali@58	2651	{
ali@58	2652	if (pswit[ECHO_SWITCH])
ali@70	2653	g_print("\n%s\n",aline);
ali@58	2654	if (!pswit[OVERVIEW_SWITCH])
ali@70	2655	g_print(" Line %ld column %ld - Spaced doublequote?\n",
ali@70	2656	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@58	2657	else
ali@58	2658	cnt_punct++;
ali@70	2659	s=g_utf8_next_char(g_utf8_next_char(t));
ali@58	2660	}
ali@92	2661	pattern=g_string_new(NULL);
ali@92	2662	for(i=0;i<G_N_ELEMENTS(single_quotes);i++)
ali@58	2663	{
ali@92	2664	g_string_assign(pattern," ");
ali@92	2665	g_string_append_unichar(pattern,single_quotes[i]);
ali@92	2666	g_string_append_c(pattern,' ');
ali@92	2667	s=aline;
ali@92	2668	while ((t=strstr(s,pattern->str)))
ali@92	2669	{
ali@92	2670	if (pswit[ECHO_SWITCH])
ali@92	2671	g_print("\n%s\n",aline);
ali@92	2672	if (!pswit[OVERVIEW_SWITCH])
ali@92	2673	g_print(" Line %ld column %ld - Spaced singlequote?\n",
ali@92	2674	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@92	2675	else
ali@92	2676	cnt_punct++;
ali@92	2677	s=g_utf8_next_char(g_utf8_next_char(t));
ali@92	2678	}
ali@58	2679	}
ali@92	2680	g_string_free(pattern,TRUE);
ali@58	2681	}
ali@58	2682
ali@58	2683	/*
ali@59	2684	* check_for_miscased_genative:
ali@59	2685	*
ali@59	2686	* Check special case of 'S instead of 's at end of word.
ali@59	2687	*/
ali@59	2688	void check_for_miscased_genative(const char *aline)
ali@59	2689	{
ali@59	2690	const char *s;
ali@70	2691	gunichar c,nc,pc;
ali@69	2692	if (!*aline)
ali@69	2693	return;
ali@70	2694	c=g_utf8_get_char(aline);
ali@70	2695	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2696	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@59	2697	{
ali@70	2698	pc=c;
ali@70	2699	c=nc;
ali@70	2700	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@92	2701	if (CHAR_IS_APOSTROPHE(c) && nc=='S' && g_unichar_islower(pc))
ali@59	2702	{
ali@59	2703	if (pswit[ECHO_SWITCH])
ali@70	2704	g_print("\n%s\n",aline);
ali@59	2705	if (!pswit[OVERVIEW_SWITCH])
ali@70	2706	g_print(" Line %ld column %ld - Capital \"S\"?\n",
ali@70	2707	linecnt,g_utf8_pointer_to_offset(aline,s)+2);
ali@59	2708	else
ali@59	2709	cnt_punct++;
ali@59	2710	}
ali@59	2711	}
ali@59	2712	}
ali@59	2713
ali@59	2714	/*
ali@60	2715	* check_end_of_line:
ali@60	2716	*
ali@60	2717	* Now check special cases - start and end of line -
ali@60	2718	* for single and double quotes. Start is sometimes [sic]
ali@60	2719	* but better to query it anyway.
ali@60	2720	* While we're here, check for dash at end of line.
ali@60	2721	*/
ali@60	2722	void check_end_of_line(const char aline,struct warnings warnings)
ali@60	2723	{
ali@70	2724	int lbytes;
ali@70	2725	const char *s;
ali@70	2726	gunichar c1,c2;
ali@70	2727	lbytes=strlen(aline);
ali@70	2728	if (g_utf8_strlen(aline,lbytes)>1)
ali@60	2729	{
ali@70	2730	s=g_utf8_prev_char(aline+lbytes);
ali@70	2731	c1=g_utf8_get_char(s);
ali@70	2732	c2=g_utf8_get_char(g_utf8_prev_char(s));
ali@94	2733	if ((CHAR_IS_DQUOTE(c1) \|\| CHAR_IS_SQUOTE(c1)) && c2==CHAR_SPACE)
ali@60	2734	{
ali@60	2735	if (pswit[ECHO_SWITCH])
ali@70	2736	g_print("\n%s\n",aline);
ali@60	2737	if (!pswit[OVERVIEW_SWITCH])
ali@70	2738	g_print(" Line %ld column %ld - Spaced quote?\n",linecnt,
ali@70	2739	g_utf8_strlen(aline,lbytes));
ali@70	2740	else
ali@70	2741	cnt_punct++;
ali@70	2742	}
ali@70	2743	c1=g_utf8_get_char(aline);
ali@70	2744	c2=g_utf8_get_char(g_utf8_next_char(aline));
ali@92	2745	if (CHAR_IS_SQUOTE(c1) && c2==CHAR_SPACE)
ali@70	2746	{
ali@70	2747	if (pswit[ECHO_SWITCH])
ali@70	2748	g_print("\n%s\n",aline);
ali@70	2749	if (!pswit[OVERVIEW_SWITCH])
ali@70	2750	g_print(" Line %ld column 1 - Spaced quote?\n",linecnt);
ali@60	2751	else
ali@60	2752	cnt_punct++;
ali@60	2753	}
ali@60	2754	/*
ali@60	2755	* Dash at end of line may well be legit - paranoid mode only
ali@60	2756	* and don't report em-dash at line-end.
ali@60	2757	*/
ali@60	2758	if (pswit[PARANOID_SWITCH] && warnings->hyphen)
ali@60	2759	{
ali@70	2760	for (s=g_utf8_prev_char(aline+lbytes);
ali@70	2761	s>aline && g_utf8_get_char(s)<=CHAR_SPACE;s=g_utf8_prev_char(s))
ali@60	2762	;
ali@70	2763	if (g_utf8_get_char(s)=='-' &&
ali@70	2764	g_utf8_get_char(g_utf8_prev_char(s))!='-')
ali@60	2765	{
ali@60	2766	if (pswit[ECHO_SWITCH])
ali@70	2767	g_print("\n%s\n",aline);
ali@60	2768	if (!pswit[OVERVIEW_SWITCH])
ali@70	2769	g_print(" Line %ld column %ld - "
ali@70	2770	"Hyphen at end of line?\n",
ali@70	2771	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@60	2772	}
ali@60	2773	}
ali@60	2774	}
ali@60	2775	}
ali@60	2776
ali@60	2777	/*
ali@61	2778	* check_for_unspaced_bracket:
ali@61	2779	*
ali@61	2780	* Brackets are often unspaced, but shouldn't be surrounded by alpha.
ali@61	2781	* If so, suspect a scanno like "a]most".
ali@61	2782	*/
ali@61	2783	void check_for_unspaced_bracket(const char *aline)
ali@61	2784	{
ali@70	2785	const char *s;
ali@70	2786	gunichar c,nc,pc;
ali@70	2787	c=g_utf8_get_char(aline);
ali@70	2788	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2789	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@61	2790	{
ali@70	2791	pc=c;
ali@70	2792	c=nc;
ali@70	2793	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@70	2794	if (!nc)
ali@70	2795	break;
ali@61	2796	/* for each bracket character in the line except 1st & last */
ali@70	2797	if (g_utf8_strchr("{[()]}",-1,c) &&
ali@70	2798	g_unichar_isalpha(pc) && g_unichar_isalpha(nc))
ali@61	2799	{
ali@61	2800	if (pswit[ECHO_SWITCH])
ali@70	2801	g_print("\n%s\n",aline);
ali@61	2802	if (!pswit[OVERVIEW_SWITCH])
ali@70	2803	g_print(" Line %ld column %ld - Unspaced bracket?\n",
ali@70	2804	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@61	2805	else
ali@61	2806	cnt_punct++;
ali@61	2807	}
ali@61	2808	}
ali@61	2809	}
ali@61	2810
ali@61	2811	/*
ali@62	2812	* check_for_unpunctuated_endquote:
ali@62	2813	*/
ali@62	2814	void check_for_unpunctuated_endquote(const char *aline)
ali@62	2815	{
ali@70	2816	const char *s;
ali@70	2817	gunichar c,nc,pc;
ali@94	2818	QuoteClass qc;
ali@70	2819	c=g_utf8_get_char(aline);
ali@70	2820	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2821	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@62	2822	{
ali@70	2823	pc=c;
ali@70	2824	c=nc;
ali@94	2825	qc=CHAR_IS_DQUOTE(c)?QUOTE_CLASS(c):INVALID_QUOTE;
ali@70	2826	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@62	2827	/* for each character in the line except 1st */
ali@97	2828	if ((qc==CLOSING_QUOTE \|\| qc==NEUTRAL_QUOTE) && g_unichar_isalpha(pc))
ali@62	2829	{
ali@62	2830	if (pswit[ECHO_SWITCH])
ali@70	2831	g_print("\n%s\n",aline);
ali@62	2832	if (!pswit[OVERVIEW_SWITCH])
ali@70	2833	g_print(" Line %ld column %ld - "
ali@70	2834	"endquote missing punctuation?\n",
ali@70	2835	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@62	2836	else
ali@62	2837	cnt_punct++;
ali@62	2838	}
ali@62	2839	}
ali@62	2840	}
ali@62	2841
ali@62	2842	/*
ali@63	2843	* check_for_html_tag:
ali@63	2844	*
ali@63	2845	* Check for <HTML TAG>.
ali@63	2846	*
ali@63	2847	* If there is a < in the line, followed at some point
ali@63	2848	* by a > then we suspect HTML.
ali@63	2849	*/
ali@63	2850	void check_for_html_tag(const char *aline)
ali@63	2851	{
ali@63	2852	const char open,close;
ali@70	2853	gchar *tag;
ali@70	2854	open=strchr(aline,'<');
ali@63	2855	if (open)
ali@63	2856	{
ali@70	2857	close=strchr(g_utf8_next_char(open),'>');
ali@63	2858	if (close)
ali@63	2859	{
ali@70	2860	if (pswit[ECHO_SWITCH])
ali@70	2861	g_print("\n%s\n",aline);
ali@70	2862	if (!pswit[OVERVIEW_SWITCH])
ali@63	2863	{
ali@70	2864	tag=g_strndup(open,close-open+1);
ali@70	2865	g_print(" Line %ld column %ld - HTML Tag? %s \n",
ali@70	2866	linecnt,g_utf8_pointer_to_offset(aline,open)+1,tag);
ali@70	2867	g_free(tag);
ali@63	2868	}
ali@70	2869	else
ali@70	2870	cnt_html++;
ali@63	2871	}
ali@63	2872	}
ali@63	2873	}
ali@63	2874
ali@63	2875	/*
ali@64	2876	* check_for_html_entity:
ali@64	2877	*
ali@64	2878	* Check for &symbol; HTML.
ali@64	2879	*
ali@64	2880	* If there is a & in the line, followed at
ali@64	2881	* some point by a ; then we suspect HTML.
ali@64	2882	*/
ali@64	2883	void check_for_html_entity(const char *aline)
ali@64	2884	{
ali@64	2885	const char s,amp,*scolon;
ali@70	2886	gchar *entity;
ali@70	2887	amp=strchr(aline,'&');
ali@64	2888	if (amp)
ali@64	2889	{
ali@70	2890	scolon=strchr(amp,';');
ali@64	2891	if (scolon)
ali@64	2892	{
ali@70	2893	for (s=amp;s<scolon;s=g_utf8_next_char(s))
ali@70	2894	if (g_utf8_get_char(s)==CHAR_SPACE)
ali@70	2895	break; /* Don't report "Jones & Son;" */
ali@70	2896	if (s>=scolon)
ali@64	2897	{
ali@64	2898	if (pswit[ECHO_SWITCH])
ali@70	2899	g_print("\n%s\n",aline);
ali@64	2900	if (!pswit[OVERVIEW_SWITCH])
ali@70	2901	{
ali@70	2902	entity=g_strndup(amp,scolon-amp+1);
ali@70	2903	g_print(" Line %ld column %d - HTML symbol? %s \n",
ali@70	2904	linecnt,(int)(amp-aline)+1,entity);
ali@70	2905	g_free(entity);
ali@70	2906	}
ali@64	2907	else
ali@64	2908	cnt_html++;
ali@64	2909	}
ali@64	2910	}
ali@64	2911	}
ali@64	2912	}
ali@64	2913
ali@65	2914	/*
ali@66	2915	* check_for_omitted_punctuation:
ali@66	2916	*
ali@66	2917	* Check for omitted punctuation at end of paragraph by working back
ali@66	2918	* through prevline. DW.
ali@66	2919	* Need to check this only for "normal" paras.
ali@66	2920	* So what is a "normal" para?
ali@66	2921	* Not normal if one-liner (chapter headings, etc.)
ali@66	2922	* Not normal if doesn't contain at least one locase letter
ali@66	2923	* Not normal if starts with space
ali@66	2924	*/
ali@66	2925	void check_for_omitted_punctuation(const char *prevline,
ali@66	2926	struct line_properties *last,int start_para_line)
ali@66	2927	{
ali@70	2928	gboolean letter_on_line=FALSE;
ali@66	2929	const char *s;
ali@92	2930	gunichar c;
ali@94	2931	gboolean closing_quote;
ali@70	2932	for (s=prevline;*s;s=g_utf8_next_char(s))
ali@70	2933	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@70	2934	{
ali@70	2935	letter_on_line=TRUE;
ali@70	2936	break;
ali@70	2937	}
ali@66	2938	/*
ali@66	2939	* This next "if" is a problem.
ali@66	2940	* If we say "start_para_line <= linecnt - 1", that includes
ali@66	2941	* one-line "paragraphs" like chapter heads. Lotsa false positives.
ali@66	2942	* If we say "start_para_line < linecnt - 1" it doesn't, but then it
ali@66	2943	* misses genuine one-line paragraphs.
ali@66	2944	*/
ali@70	2945	if (letter_on_line && last->blen>2 && start_para_line<linecnt-1 &&
ali@70	2946	g_utf8_get_char(prevline)>CHAR_SPACE)
ali@66	2947	{
ali@92	2948	s=prevline+strlen(prevline);
ali@92	2949	do
ali@92	2950	{
ali@92	2951	s=g_utf8_prev_char(s);
ali@92	2952	c=g_utf8_get_char(s);
ali@94	2953	if (QUOTE_CLASS(c)==CLOSING_QUOTE \|\| QUOTE_CLASS(c)==NEUTRAL_QUOTE)
ali@94	2954	closing_quote=TRUE;
ali@94	2955	else
ali@94	2956	closing_quote=FALSE;
ali@94	2957	} while (closing_quote && s>prevline);
ali@70	2958	for (;s>prevline;s=g_utf8_prev_char(s))
ali@66	2959	{
ali@70	2960	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@66	2961	{
ali@66	2962	if (pswit[ECHO_SWITCH])
ali@70	2963	g_print("\n%s\n",prevline);
ali@66	2964	if (!pswit[OVERVIEW_SWITCH])
ali@70	2965	g_print(" Line %ld column %ld - "
ali@66	2966	"No punctuation at para end?\n",
ali@70	2967	linecnt-1,g_utf8_strlen(prevline,-1));
ali@66	2968	else
ali@66	2969	cnt_punct++;
ali@66	2970	break;
ali@66	2971	}
ali@97	2972	if (g_utf8_strchr("—-.:!([{?}])",-1,g_utf8_get_char(s)))
ali@66	2973	break;
ali@66	2974	}
ali@66	2975	}
ali@66	2976	}
ali@66	2977
ali@69	2978	gboolean report_duplicate_queries(gpointer key,gpointer value,gpointer data)
ali@69	2979	{
ali@69	2980	const char *word=key;
ali@69	2981	int *dupcnt=value;
ali@69	2982	if (*dupcnt)
ali@70	2983	g_print("\nNote: Queried word %s was duplicated %d times\n",
ali@69	2984	word,*dupcnt);
ali@69	2985	return FALSE;
ali@69	2986	}
ali@69	2987
ali@70	2988	void print_as_windows_1252(const char *string)
ali@70	2989	{
ali@70	2990	gsize inbytes,outbytes;
ali@70	2991	gchar buf,bp;
ali@86	2992	static GIConv converter=(GIConv)-1;
ali@70	2993	if (!string)
ali@70	2994	{
ali@70	2995	if (converter!=(GIConv)-1)
ali@70	2996	g_iconv_close(converter);
ali@70	2997	converter=(GIConv)-1;
ali@70	2998	return;
ali@70	2999	}
ali@86	3000	if (converter==(GIConv)-1)
ali@70	3001	converter=g_iconv_open("WINDOWS-1252","UTF-8");
ali@70	3002	if (converter!=(GIConv)-1)
ali@70	3003	{
ali@70	3004	inbytes=outbytes=strlen(string);
ali@70	3005	bp=buf=g_malloc(outbytes+1);
ali@70	3006	g_iconv(converter,(char **)&string,&inbytes,&bp,&outbytes);
ali@70	3007	*bp='\0';
ali@70	3008	fputs(buf,stdout);
ali@70	3009	g_free(buf);
ali@70	3010	}
ali@70	3011	else
ali@70	3012	fputs(string,stdout);
ali@70	3013	}
ali@70	3014
ali@72	3015	void print_as_utf_8(const char *string)
ali@72	3016	{
ali@72	3017	fputs(string,stdout);
ali@72	3018	}
ali@72	3019
ali@66	3020	/*
ali@41	3021	* procfile:
ali@41	3022	*
ali@41	3023	* Process one file.
ali@41	3024	*/
ali@69	3025	void procfile(const char *filename)
ali@41	3026	{
ali@65	3027	const char *s;
ali@69	3028	gchar parastart=NULL; / first line of current para */
ali@69	3029	gchar etext,aline;
ali@69	3030	gchar *etext_ptr;
ali@69	3031	GError *err=NULL;
ali@41	3032	struct first_pass_results *first_pass_results;
ali@42	3033	struct warnings *warnings;
ali@43	3034	struct counters counters={0};
ali@45	3035	struct line_properties last={0};
ali@56	3036	struct parities parities={0};
ali@69	3037	struct pending pending={0};
ali@69	3038	gboolean isemptyline;
ali@68	3039	long start_para_line=0;
ali@69	3040	gboolean isnewpara=FALSE,enddash=FALSE;
ali@45	3041	last.start=CHAR_SPACE;
ali@68	3042	linecnt=checked_linecnt=0;
ali@69	3043	etext=read_etext(filename,&err);
ali@69	3044	if (!etext)
ali@41	3045	{
ali@68	3046	if (pswit[STDOUT_SWITCH])
ali@69	3047	fprintf(stdout,"bookloupe: %s: %s\n",filename,err->message);
ali@68	3048	else
ali@69	3049	fprintf(stderr,"bookloupe: %s: %s\n",filename,err->message);
ali@41	3050	exit(1);
ali@41	3051	}
ali@70	3052	g_print("\n\nFile: %s\n\n",filename);
ali@69	3053	first_pass_results=first_pass(etext);
ali@42	3054	warnings=report_first_pass(first_pass_results);
ali@69	3055	qword=g_tree_new_full((GCompareDataFunc)strcmp,NULL,g_free,g_free);
ali@69	3056	qperiod=g_tree_new_full((GCompareDataFunc)strcmp,NULL,g_free,NULL);
ali@40	3057	/*
ali@40	3058	* Here we go with the main pass. Hold onto yer hat!
ali@40	3059	*/
ali@65	3060	linecnt=0;
ali@69	3061	etext_ptr=etext;
ali@101	3062	while ((aline=flgets(&etext_ptr,linecnt+1,warnings->newlines)))
ali@40	3063	{
ali@68	3064	linecnt++;
ali@68	3065	if (linecnt==1)
ali@69	3066	isnewpara=TRUE;
ali@70	3067	if (pswit[DP_SWITCH] && g_str_has_prefix(aline,"-----File: "))
ali@40	3068	continue; // skip DP page separators completely
ali@68	3069	if (linecnt<first_pass_results->firstline \|\|
ali@41	3070	(first_pass_results->footerline>0 &&
ali@41	3071	linecnt>first_pass_results->footerline))
ali@40	3072	{
ali@68	3073	if (pswit[HEADER_SWITCH])
ali@40	3074	{
ali@70	3075	if (g_str_has_prefix(aline,"Title:"))
ali@70	3076	g_print(" %s\n",aline);
ali@70	3077	if (g_str_has_prefix(aline,"Author:"))
ali@70	3078	g_print(" %s\n",aline);
ali@70	3079	if (g_str_has_prefix(aline,"Release Date:"))
ali@70	3080	g_print(" %s\n",aline);
ali@70	3081	if (g_str_has_prefix(aline,"Edition:"))
ali@70	3082	g_print(" %s\n\n",aline);
ali@40	3083	}
ali@68	3084	continue; /* skip through the header */
ali@40	3085	}
ali@68	3086	checked_linecnt++;
ali@65	3087	print_pending(aline,parastart,&pending);
ali@98	3088	isemptyline=analyse_quotes(aline,&counters);
ali@68	3089	if (isnewpara && !isemptyline)
ali@40	3090	{
ali@40	3091	/* This line is the start of a new paragraph. */
ali@68	3092	start_para_line=linecnt;
ali@40	3093	/* Capture its first line in case we want to report it later. */
ali@69	3094	g_free(parastart);
ali@69	3095	parastart=g_strdup(aline);
ali@56	3096	memset(&parities,0,sizeof(parities)); /* restart the quote count */
ali@68	3097	s=aline;
ali@70	3098	while (*s && !g_unichar_isalpha(g_utf8_get_char(s)) &&
ali@70	3099	!g_unichar_isdigit(g_utf8_get_char(s)))
ali@70	3100	s=g_utf8_next_char(s);
ali@70	3101	if (g_unichar_islower(g_utf8_get_char(s)))
ali@40	3102	{
ali@40	3103	/* and its first letter is lowercase */
ali@68	3104	if (pswit[ECHO_SWITCH])
ali@70	3105	g_print("\n%s\n",aline);
ali@68	3106	if (!pswit[OVERVIEW_SWITCH])
ali@70	3107	g_print(" Line %ld column %ld - "
ali@40	3108	"Paragraph starts with lower-case\n",
ali@70	3109	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@68	3110	else
ali@68	3111	cnt_punct++;
ali@40	3112	}
ali@69	3113	isnewpara=FALSE; /* Signal the end of new para processing. */
ali@40	3114	}
ali@68	3115	/* Check for an em-dash broken at line end. */
ali@70	3116	if (enddash && g_utf8_get_char(aline)=='-')
ali@40	3117	{
ali@68	3118	if (pswit[ECHO_SWITCH])
ali@70	3119	g_print("\n%s\n",aline);
ali@68	3120	if (!pswit[OVERVIEW_SWITCH])
ali@70	3121	g_print(" Line %ld column 1 - Broken em-dash?\n",linecnt);
ali@68	3122	else
ali@68	3123	cnt_punct++;
ali@40	3124	}
ali@69	3125	enddash=FALSE;
ali@70	3126	for (s=g_utf8_prev_char(aline+strlen(aline));
ali@70	3127	g_utf8_get_char(s)==' ' && s>aline;s=g_utf8_prev_char(s))
ali@40	3128	;
ali@70	3129	if (s>=aline && g_utf8_get_char(s)=='-')
ali@69	3130	enddash=TRUE;
ali@67	3131	check_for_control_characters(aline);
ali@103	3132	check_for_odd_characters(aline,warnings,isemptyline);
ali@68	3133	if (warnings->longline)
ali@45	3134	check_for_long_line(aline);
ali@68	3135	if (warnings->shortline)
ali@45	3136	check_for_short_line(aline,&last);
ali@68	3137	last.blen=last.len;
ali@70	3138	last.len=g_utf8_strlen(aline,-1);
ali@70	3139	last.start=g_utf8_get_char(aline);
ali@46	3140	check_for_starting_punctuation(aline);
ali@68	3141	if (warnings->dash)
ali@40	3142	{
ali@47	3143	check_for_spaced_emdash(aline);
ali@47	3144	check_for_spaced_dash(aline);
ali@40	3145	}
ali@48	3146	check_for_unmarked_paragraphs(aline);
ali@49	3147	check_for_jeebies(aline);
ali@50	3148	check_for_mta_from(aline);
ali@51	3149	check_for_orphan_character(aline);
ali@52	3150	check_for_pling_scanno(aline);
ali@53	3151	check_for_extra_period(aline,warnings);
ali@54	3152	check_for_following_punctuation(aline);
ali@55	3153	check_for_typos(aline,warnings);
ali@56	3154	check_for_misspaced_punctuation(aline,&parities,isemptyline);
ali@57	3155	check_for_double_punctuation(aline,warnings);
ali@58	3156	check_for_spaced_quotes(aline);
ali@59	3157	check_for_miscased_genative(aline);
ali@60	3158	check_end_of_line(aline,warnings);
ali@61	3159	check_for_unspaced_bracket(aline);
ali@68	3160	if (warnings->endquote)
ali@62	3161	check_for_unpunctuated_endquote(aline);
ali@63	3162	check_for_html_tag(aline);
ali@64	3163	check_for_html_entity(aline);
ali@68	3164	if (isemptyline)
ali@40	3165	{
ali@65	3166	check_for_mismatched_quotes(&counters,&pending);
ali@93	3167	counters_reset(&counters);
ali@40	3168	/* let the next iteration know that it's starting a new para */
ali@69	3169	isnewpara=TRUE;
ali@69	3170	if (prevline)
ali@69	3171	check_for_omitted_punctuation(prevline,&last,start_para_line);
ali@40	3172	}
ali@69	3173	g_free(prevline);
ali@69	3174	prevline=g_strdup(aline);
ali@0	3175	}
ali@93	3176	linecnt++;
ali@93	3177	check_for_mismatched_quotes(&counters,&pending);
ali@93	3178	print_pending(NULL,parastart,&pending);
ali@93	3179	reset_pending(&pending);
ali@69	3180	if (prevline)
ali@69	3181	{
ali@69	3182	g_free(prevline);
ali@69	3183	prevline=NULL;
ali@69	3184	}
ali@69	3185	g_free(parastart);
ali@69	3186	g_free(prevline);
ali@69	3187	g_free(etext);
ali@79	3188	if (!pswit[OVERVIEW_SWITCH] && !pswit[VERBOSE_SWITCH])
ali@69	3189	g_tree_foreach(qword,report_duplicate_queries,NULL);
ali@69	3190	g_tree_unref(qword);
ali@69	3191	g_tree_unref(qperiod);
ali@92	3192	counters_destroy(&counters);
ali@70	3193	g_set_print_handler(NULL);
ali@70	3194	print_as_windows_1252(NULL);
ali@71	3195	if (pswit[MARKUP_SWITCH])
ali@71	3196	loseentities(NULL);
ali@0	3197	}
ali@0	3198
ali@40	3199	/*
ali@40	3200	* flgets:
ali@40	3201	*
ali@101	3202	* Get one line from the input text. The setting of newlines has the following
ali@101	3203	* effect:
ali@101	3204	*
ali@101	3205	* DOS_NEWLINES: Check for the existence of exactly one CR-LF line-end per line.
ali@101	3206	*
ali@101	3207	* OS9_NEWLINES: Asserts that etext contains no LFs. CR is used as
ali@101	3208	* the newline character.
ali@101	3209	*
ali@101	3210	* UNIX_NEWLINES: Check for the presence of CRs.
ali@101	3211	*
ali@101	3212	* In all cases, check that the last line is correctly terminated.
ali@40	3213	*
ali@40	3214	* Returns: a pointer to the line.
ali@40	3215	*/
ali@101	3216	char flgets(char *etext,long lcnt,int newlines)
ali@0	3217	{
ali@70	3218	gunichar c;
ali@69	3219	gboolean isCR=FALSE;
ali@69	3220	char theline=etext;
ali@70	3221	char *eos=theline;
ali@70	3222	gchar *s;
ali@70	3223	for (;;)
ali@40	3224	{
ali@70	3225	c=g_utf8_get_char(*etext);
ali@99	3226	if (!c)
ali@99	3227	{
ali@99	3228	if (*etext==theline)
ali@99	3229	return NULL;
ali@99	3230	else if (pswit[LINE_END_SWITCH])
ali@99	3231	{
ali@99	3232	if (pswit[ECHO_SWITCH])
ali@99	3233	{
ali@99	3234	s=g_strndup(theline,eos-theline);
ali@99	3235	g_print("\n%s\n",s);
ali@99	3236	g_free(s);
ali@99	3237	}
ali@99	3238	if (!pswit[OVERVIEW_SWITCH])
ali@101	3239	{
ali@101	3240	if (newlines==OS9_NEWLINES)
ali@101	3241	g_print(" Line %ld - No CR?\n",lcnt);
ali@101	3242	else
ali@101	3243	{
ali@101	3244	/* There may, or may not, have been a CR */
ali@101	3245	g_print(" Line %ld - No LF?\n",lcnt);
ali@101	3246	}
ali@101	3247	}
ali@99	3248	else
ali@99	3249	cnt_lineend++;
ali@99	3250	}
ali@99	3251	break;
ali@99	3252	}
ali@70	3253	etext=g_utf8_next_char(etext);
ali@40	3254	/* either way, it's end of line */
ali@69	3255	if (c=='\n')
ali@40	3256	{
ali@101	3257	if (newlines==DOS_NEWLINES && !isCR)
ali@40	3258	{
ali@40	3259	/* Error - a LF without a preceding CR */
ali@68	3260	if (pswit[LINE_END_SWITCH])
ali@40	3261	{
ali@68	3262	if (pswit[ECHO_SWITCH])
ali@70	3263	{
ali@70	3264	s=g_strndup(theline,eos-theline);
ali@70	3265	g_print("\n%s\n",s);
ali@70	3266	g_free(s);
ali@70	3267	}
ali@68	3268	if (!pswit[OVERVIEW_SWITCH])
ali@70	3269	g_print(" Line %ld - No CR?\n",lcnt);
ali@68	3270	else
ali@68	3271	cnt_lineend++;
ali@40	3272	}
ali@40	3273	}
ali@101	3274	break;
ali@40	3275	}
ali@69	3276	if (c=='\r')
ali@40	3277	{
ali@101	3278	if (newlines==OS9_NEWLINES)
ali@101	3279	break;
ali@101	3280	if (isCR \|\| newlines==UNIX_NEWLINES)
ali@40	3281	{
ali@68	3282	if (pswit[LINE_END_SWITCH])
ali@40	3283	{
ali@68	3284	if (pswit[ECHO_SWITCH])
ali@70	3285	{
ali@70	3286	s=g_strndup(theline,eos-theline);
ali@70	3287	g_print("\n%s\n",s);
ali@70	3288	g_free(s);
ali@70	3289	}
ali@68	3290	if (!pswit[OVERVIEW_SWITCH])
ali@101	3291	{
ali@101	3292	if (newlines==UNIX_NEWLINES)
ali@101	3293	g_print(" Line %ld column %ld - Embedded CR?\n",
ali@101	3294	lcnt,g_utf8_pointer_to_offset(theline,eos)+1);
ali@101	3295	else
ali@101	3296	g_print(" Line %ld - Two successive CRs?\n",
ali@101	3297	lcnt);
ali@101	3298	}
ali@68	3299	else
ali@68	3300	cnt_lineend++;
ali@40	3301	}
ali@101	3302	if (newlines==UNIX_NEWLINES)
ali@101	3303	*eos=' ';
ali@40	3304	}
ali@101	3305	if (newlines==DOS_NEWLINES)
ali@101	3306	isCR=TRUE;
ali@40	3307	}
ali@68	3308	else
ali@40	3309	{
ali@68	3310	if (pswit[LINE_END_SWITCH] && isCR)
ali@40	3311	{
ali@68	3312	if (pswit[ECHO_SWITCH])
ali@70	3313	{
ali@70	3314	s=g_strndup(theline,eos-theline);
ali@70	3315	g_print("\n%s\n",s);
ali@70	3316	g_free(s);
ali@70	3317	}
ali@68	3318	if (!pswit[OVERVIEW_SWITCH])
ali@70	3319	g_print(" Line %ld column %ld - CR without LF?\n",
ali@70	3320	lcnt,g_utf8_pointer_to_offset(theline,eos)+1);
ali@68	3321	else
ali@68	3322	cnt_lineend++;
ali@70	3323	*eos=' ';
ali@40	3324	}
ali@69	3325	isCR=FALSE;
ali@70	3326	eos=g_utf8_next_char(eos);
ali@40	3327	}
ali@69	3328	}
ali@70	3329	*eos='\0';
ali@0	3330	if (pswit[MARKUP_SWITCH])
ali@68	3331	postprocess_for_HTML(theline);
ali@0	3332	if (pswit[DP_SWITCH])
ali@68	3333	postprocess_for_DP(theline);
ali@40	3334	return theline;
ali@0	3335	}
ali@0	3336
ali@40	3337	/*
ali@40	3338	* mixdigit:
ali@40	3339	*
ali@40	3340	* Takes a "word" as a parameter, and checks whether it
ali@40	3341	* contains a mixture of alpha and digits. Generally, this is an
ali@40	3342	* error, but may not be for cases like 4th or L5 12s. 3d.
ali@40	3343	*
ali@70	3344	* Returns: TRUE iff an is error found.
ali@40	3345	*/
ali@70	3346	gboolean mixdigit(const char *checkword)
ali@0	3347	{
ali@70	3348	gboolean wehaveadigit,wehavealetter,query;
ali@70	3349	const char s,nondigit;
ali@70	3350	wehaveadigit=wehavealetter=query=FALSE;
ali@70	3351	for (s=checkword;*s;s=g_utf8_next_char(s))
ali@70	3352	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@70	3353	wehavealetter=TRUE;
ali@70	3354	else if (g_unichar_isdigit(g_utf8_get_char(s)))
ali@70	3355	wehaveadigit=TRUE;
ali@40	3356	if (wehaveadigit && wehavealetter)
ali@40	3357	{
ali@40	3358	/* Now exclude common legit cases, like "21st" and "12l. 3s. 11d." */
ali@70	3359	query=TRUE;
ali@70	3360	for (nondigit=checkword;g_unichar_isdigit(g_utf8_get_char(nondigit));
ali@70	3361	nondigit=g_utf8_next_char(nondigit))
ali@68	3362	;
ali@68	3363	/* digits, ending in st, rd, nd, th of either case */
ali@70	3364	if (!g_ascii_strcasecmp(nondigit,"st") \|\|
ali@70	3365	!g_ascii_strcasecmp(nondigit,"rd") \|\|
ali@70	3366	!g_ascii_strcasecmp(nondigit,"nd") \|\|
ali@70	3367	!g_ascii_strcasecmp(nondigit,"th"))
ali@70	3368	query=FALSE;
ali@70	3369	if (!g_ascii_strcasecmp(nondigit,"sts") \|\|
ali@70	3370	!g_ascii_strcasecmp(nondigit,"rds") \|\|
ali@70	3371	!g_ascii_strcasecmp(nondigit,"nds") \|\|
ali@70	3372	!g_ascii_strcasecmp(nondigit,"ths"))
ali@70	3373	query=FALSE;
ali@70	3374	if (!g_ascii_strcasecmp(nondigit,"stly") \|\|
ali@70	3375	!g_ascii_strcasecmp(nondigit,"rdly") \|\|
ali@70	3376	!g_ascii_strcasecmp(nondigit,"ndly") \|\|
ali@70	3377	!g_ascii_strcasecmp(nondigit,"thly"))
ali@70	3378	query=FALSE;
ali@68	3379	/* digits, ending in l, L, s or d */
ali@70	3380	if (!g_ascii_strcasecmp(nondigit,"l") \|\| !strcmp(nondigit,"s") \|\|
ali@70	3381	!strcmp(nondigit,"d"))
ali@70	3382	query=FALSE;
ali@68	3383	/*
ali@40	3384	* L at the start of a number, representing Britsh pounds, like L500.
ali@70	3385	* This is cute. We know the current word is mixed digit. If the first
ali@68	3386	* letter is L, there must be at least one digit following. If both
ali@68	3387	* digits and letters follow, we have a genuine error, else we have a
ali@68	3388	* capital L followed by digits, and we accept that as a non-error.
ali@40	3389	*/
ali@70	3390	if (g_utf8_get_char(checkword)=='L' &&
ali@70	3391	!mixdigit(g_utf8_next_char(checkword)))
ali@70	3392	query=FALSE;
ali@40	3393	}
ali@40	3394	return query;
ali@0	3395	}
ali@0	3396
ali@40	3397	/*
ali@40	3398	* getaword:
ali@40	3399	*
ali@69	3400	* Extracts the first/next "word" from the line, and returns it.
ali@69	3401	* A word is defined as one English word unit--or at least that's the aim.
ali@69	3402	* "ptr" is advanced to the position in the line where we will start
ali@69	3403	* looking for the next word.
ali@40	3404	*
ali@69	3405	* Returns: A newly-allocated string.
ali@40	3406	*/
ali@69	3407	gchar getaword(const char *ptr)
ali@0	3408	{
ali@70	3409	const char s,t;
ali@69	3410	GString *word;
ali@70	3411	gunichar c,pc;
ali@69	3412	word=g_string_new(NULL);
ali@70	3413	for (;!g_unichar_isdigit(g_utf8_get_char(*ptr)) &&
ali@70	3414	!g_unichar_isalpha(g_utf8_get_char(*ptr)) &&
ali@70	3415	*ptr;ptr=g_utf8_next_char(*ptr))
ali@100	3416	{
ali@100	3417	/* Handle exceptions for footnote markers like [1] */
ali@100	3418	if (g_utf8_get_char(*ptr)=='[')
ali@100	3419	{
ali@100	3420	g_string_append_c(word,'[');
ali@100	3421	s=g_utf8_next_char(*ptr);
ali@100	3422	for (;g_unichar_isdigit(g_utf8_get_char(s));s=g_utf8_next_char(s))
ali@100	3423	g_string_append_unichar(word,g_utf8_get_char(s));
ali@100	3424	if (g_utf8_get_char(s)==']')
ali@100	3425	{
ali@100	3426	g_string_append_c(word,']');
ali@100	3427	*ptr=g_utf8_next_char(s);
ali@100	3428	return g_string_free(word,FALSE);
ali@100	3429	}
ali@100	3430	else
ali@100	3431	g_string_truncate(word,0);
ali@100	3432	}
ali@100	3433	}
ali@40	3434	/*
ali@40	3435	* Use a look-ahead to handle exceptions for numbers like 1,000 and 1.35.
ali@40	3436	* Especially yucky is the case of L1,000
ali@40	3437	* This section looks for a pattern of characters including a digit
ali@40	3438	* followed by a comma or period followed by one or more digits.
ali@40	3439	* If found, it returns this whole pattern as a word; otherwise we discard
ali@40	3440	* the results and resume our normal programming.
ali@40	3441	*/
ali@69	3442	s=*ptr;
ali@70	3443	for (;g_unichar_isdigit(g_utf8_get_char(s)) \|\|
ali@70	3444	g_unichar_isalpha(g_utf8_get_char(s)) \|\|
ali@70	3445	g_utf8_get_char(s)==',' \|\| g_utf8_get_char(s)=='.';s=g_utf8_next_char(s))
ali@70	3446	g_string_append_unichar(word,g_utf8_get_char(s));
ali@82	3447	if (word->len)
ali@40	3448	{
ali@82	3449	for (t=g_utf8_next_char(word->str);*t;t=g_utf8_next_char(t))
ali@40	3450	{
ali@82	3451	c=g_utf8_get_char(t);
ali@82	3452	pc=g_utf8_get_char(g_utf8_prev_char(t));
ali@82	3453	if ((c=='.' \|\| c==',') && g_unichar_isdigit(pc))
ali@82	3454	{
ali@82	3455	*ptr=s;
ali@82	3456	return g_string_free(word,FALSE);
ali@82	3457	}
ali@40	3458	}
ali@40	3459	}
ali@0	3460	/* we didn't find a punctuated number - do the regular getword thing */
ali@69	3461	g_string_truncate(word,0);
ali@92	3462	c=g_utf8_get_char(*ptr);
ali@92	3463	for (;g_unichar_isdigit(c) \|\| g_unichar_isalpha(c) \|\| CHAR_IS_APOSTROPHE(c);
ali@92	3464	ptr=g_utf8_next_char(ptr),c=g_utf8_get_char(*ptr))
ali@92	3465	g_string_append_unichar(word,c);
ali@69	3466	return g_string_free(word,FALSE);
ali@0	3467	}
ali@0	3468
ali@40	3469	/*
ali@40	3470	* isroman:
ali@40	3471	*
ali@40	3472	* Is this word a Roman Numeral?
ali@40	3473	*
ali@40	3474	* It doesn't actually validate that the number is a valid Roman Numeral--for
ali@40	3475	* example it will pass MXXXXXXXXXX as a valid Roman Numeral, but that's not
ali@40	3476	* what we're here to do. If it passes this, it LOOKS like a Roman numeral.
ali@40	3477	* Anyway, the actual Romans were pretty tolerant of bad arithmetic, or
ali@40	3478	* expressions thereof, except when it came to taxes. Allow any number of M,
ali@40	3479	* an optional D, an optional CM or CD, any number of optional Cs, an optional
ali@40	3480	* XL or an optional XC, an optional IX or IV, an optional V and any number
ali@40	3481	* of optional Is.
ali@40	3482	*/
ali@69	3483	gboolean isroman(const char *t)
ali@0	3484	{
ali@69	3485	const char *s;
ali@40	3486	if (!t \|\| !*t)
ali@69	3487	return FALSE;
ali@40	3488	s=t;
ali@70	3489	while (g_utf8_get_char(t)=='m' && *t)
ali@40	3490	t++;
ali@70	3491	if (g_utf8_get_char(t)=='d')
ali@40	3492	t++;
ali@70	3493	if (g_str_has_prefix(t,"cm"))
ali@40	3494	t+=2;
ali@70	3495	if (g_str_has_prefix(t,"cd"))
ali@40	3496	t+=2;
ali@70	3497	while (g_utf8_get_char(t)=='c' && *t)
ali@40	3498	t++;
ali@70	3499	if (g_str_has_prefix(t,"xl"))
ali@40	3500	t+=2;
ali@70	3501	if (g_str_has_prefix(t,"xc"))
ali@40	3502	t+=2;
ali@70	3503	if (g_utf8_get_char(t)=='l')
ali@40	3504	t++;
ali@70	3505	while (g_utf8_get_char(t)=='x' && *t)
ali@40	3506	t++;
ali@70	3507	if (g_str_has_prefix(t,"ix"))
ali@40	3508	t+=2;
ali@70	3509	if (g_str_has_prefix(t,"iv"))
ali@40	3510	t+=2;
ali@70	3511	if (g_utf8_get_char(t)=='v')
ali@40	3512	t++;
ali@70	3513	while (g_utf8_get_char(t)=='i' && *t)
ali@40	3514	t++;
ali@40	3515	return !*t;
ali@0	3516	}
ali@0	3517
ali@40	3518	/*
ali@40	3519	* postprocess_for_DP:
ali@40	3520	*
ali@40	3521	* Invoked with the -d switch from flgets().
ali@40	3522	* It simply "removes" from the line a hard-coded set of common
ali@40	3523	* DP-specific tags, so that the line passed to the main routine has
ali@40	3524	* been pre-cleaned of DP markup.
ali@40	3525	*/
ali@0	3526	void postprocess_for_DP(char *theline)
ali@0	3527	{
ali@40	3528	char s,t;
ali@0	3529	int i;
ali@0	3530	if (!*theline)
ali@68	3531	return;
ali@40	3532	for (i=0;*DPmarkup[i];i++)
ali@70	3533	while ((s=strstr(theline,DPmarkup[i])))
ali@40	3534	{
ali@68	3535	t=s+strlen(DPmarkup[i]);
ali@70	3536	memmove(s,t,strlen(t)+1);
ali@40	3537	}
ali@0	3538	}
ali@0	3539
ali@40	3540	/*
ali@40	3541	* postprocess_for_HTML:
ali@40	3542	*
ali@40	3543	* Invoked with the -m switch from flgets().
ali@40	3544	* It simply "removes" from the line a hard-coded set of common
ali@40	3545	* HTML tags and "replaces" a hard-coded set of common HTML
ali@40	3546	* entities, so that the line passed to the main routine has
ali@40	3547	* been pre-cleaned of HTML.
ali@40	3548	*/
ali@0	3549	void postprocess_for_HTML(char *theline)
ali@0	3550	{
ali@70	3551	while (losemarkup(theline))
ali@70	3552	;
ali@71	3553	loseentities(theline);
ali@0	3554	}
ali@0	3555
ali@0	3556	char losemarkup(char theline)
ali@0	3557	{
ali@40	3558	char s,t;
ali@0	3559	int i;
ali@70	3560	s=strchr(theline,'<');
ali@70	3561	t=s?strchr(s,'>'):NULL;
ali@40	3562	if (!s \|\| !t)
ali@40	3563	return NULL;
ali@40	3564	for (i=0;*markup[i];i++)
ali@70	3565	if (tagcomp(g_utf8_next_char(s),markup[i]))
ali@40	3566	{
ali@70	3567	t=g_utf8_next_char(t);
ali@70	3568	memmove(s,t,strlen(t)+1);
ali@70	3569	return s;
ali@68	3570	}
ali@40	3571	/* It's an unrecognized <xxx>. */
ali@40	3572	return NULL;
ali@0	3573	}
ali@0	3574
ali@71	3575	void loseentities(char *theline)
ali@0	3576	{
ali@0	3577	int i;
ali@71	3578	gsize nb;
ali@71	3579	char amp,scolon;
ali@71	3580	gchar s,t;
ali@71	3581	gunichar c;
ali@71	3582	GTree *entities=NULL;
ali@86	3583	static GIConv translit=(GIConv)-1,to_utf8=(GIConv)-1;
ali@71	3584	if (!theline)
ali@40	3585	{
ali@71	3586	if (entities)
ali@71	3587	g_tree_destroy(entities);
ali@71	3588	entities=NULL;
ali@86	3589	if (translit!=(GIConv)-1)
ali@71	3590	g_iconv_close(translit);
ali@71	3591	translit=(GIConv)-1;
ali@86	3592	if (to_utf8!=(GIConv)-1)
ali@71	3593	g_iconv_close(to_utf8);
ali@71	3594	to_utf8=(GIConv)-1;
ali@71	3595	return;
ali@71	3596	}
ali@71	3597	if (!*theline)
ali@71	3598	return;
ali@71	3599	if (!entities)
ali@71	3600	{
ali@71	3601	entities=g_tree_new((GCompareFunc)strcmp);
ali@71	3602	for(i=0;i<G_N_ELEMENTS(HTMLentities);i++)
ali@71	3603	g_tree_insert(entities,HTMLentities[i].name,
ali@71	3604	GUINT_TO_POINTER(HTMLentities[i].c));
ali@71	3605	}
ali@71	3606	if (translit==(GIConv)-1)
ali@71	3607	translit=g_iconv_open("ISO_8859-1//TRANSLIT","UTF-8");
ali@71	3608	if (to_utf8==(GIConv)-1)
ali@71	3609	to_utf8=g_iconv_open("UTF-8","ISO_8859-1");
ali@71	3610	while((amp=strchr(theline,'&')))
ali@71	3611	{
ali@71	3612	scolon=strchr(amp,';');
ali@71	3613	if (scolon)
ali@40	3614	{
ali@71	3615	if (amp[1]=='#')
ali@71	3616	{
ali@71	3617	if (amp+2+strspn(amp+2,"0123456789")==scolon)
ali@71	3618	c=strtol(amp+2,NULL,10);
ali@71	3619	else if (amp[2]=='x' &&
ali@71	3620	amp+3+strspn(amp+3,"0123456789abcdefABCDEF")==scolon)
ali@71	3621	c=strtol(amp+3,NULL,16);
ali@71	3622	}
ali@71	3623	else
ali@71	3624	{
ali@71	3625	s=g_strndup(amp+1,scolon-(amp+1));
ali@71	3626	c=GPOINTER_TO_UINT(g_tree_lookup(entities,s));
ali@71	3627	g_free(s);
ali@71	3628	}
ali@40	3629	}
ali@71	3630	else
ali@71	3631	c=0;
ali@71	3632	if (c)
ali@71	3633	{
ali@71	3634	theline=amp;
ali@71	3635	if (c<128 \|\| c>=192 && c<=255) /* An ISO-8859-1 character */
ali@71	3636	theline+=g_unichar_to_utf8(c,theline);
ali@71	3637	else
ali@71	3638	{
ali@71	3639	s=g_malloc(6);
ali@71	3640	nb=g_unichar_to_utf8(c,s);
ali@71	3641	t=g_convert_with_iconv(s,nb,translit,NULL,&nb,NULL);
ali@71	3642	g_free(s);
ali@71	3643	s=g_convert_with_iconv(t,nb,to_utf8,NULL,&nb,NULL);
ali@71	3644	g_free(t);
ali@71	3645	memcpy(theline,s,nb);
ali@71	3646	g_free(s);
ali@71	3647	theline+=nb;
ali@71	3648	}
ali@71	3649	memmove(theline,g_utf8_next_char(scolon),
ali@71	3650	strlen(g_utf8_next_char(scolon))+1);
ali@71	3651	}
ali@71	3652	else
ali@71	3653	theline=g_utf8_next_char(amp);
ali@40	3654	}
ali@0	3655	}
ali@0	3656
ali@70	3657	gboolean tagcomp(const char strin,const char basetag)
ali@0	3658	{
ali@70	3659	gboolean retval;
ali@70	3660	gchar s,t;
ali@70	3661	if (g_utf8_get_char(strin)=='/')
ali@70	3662	t=g_utf8_casefold(g_utf8_next_char(strin),-1); /* ignore a slash */
ali@70	3663	else
ali@70	3664	t=g_utf8_casefold(strin,-1);
ali@70	3665	s=g_utf8_casefold(basetag,-1);
ali@70	3666	retval=g_str_has_prefix(t,s);
ali@70	3667	g_free(s);
ali@70	3668	g_free(t);
ali@70	3669	return retval;
ali@0	3670	}
ali@0	3671
ali@69	3672	void proghelp(GOptionContext *context)
ali@0	3673	{
ali@69	3674	gchar *help;
ali@40	3675	fputs("Bookloupe version " PACKAGE_VERSION ".\n",stderr);
ali@40	3676	fputs("Copyright 2000-2005 Jim Tinsley <jtinsley@pobox.com>.\n",stderr);
ali@40	3677	fputs("Copyright 2012- J. Ali Harlow <ali@juiblex.co.uk>.\n",stderr);
ali@40	3678	fputs("Bookloupe comes wih ABSOLUTELY NO WARRANTY. "
ali@40	3679	"For details, read the file COPYING.\n",stderr);
ali@40	3680	fputs("This is Free Software; "
ali@40	3681	"you may redistribute it under certain conditions (GPL);\n",stderr);
ali@40	3682	fputs("read the file COPYING for details.\n\n",stderr);
ali@69	3683	help=g_option_context_get_help(context,TRUE,NULL);
ali@69	3684	fputs(help,stderr);
ali@69	3685	g_free(help);
ali@69	3686	fputs("Sample usage: bookloupe warpeace.txt\n\n",stderr);
ali@40	3687	fputs("Bookloupe queries anything it thinks shouldn't be in a PG text; "
ali@40	3688	"non-ASCII\n",stderr);
ali@40	3689	fputs("characters like accented letters, "
ali@40	3690	"lines longer than 75 or shorter than 55,\n",stderr);
ali@40	3691	fputs("unbalanced quotes or brackets, "
ali@40	3692	"a variety of badly formatted punctuation, \n",stderr);
ali@40	3693	fputs("HTML tags, some likely typos. "
ali@40	3694	"It is NOT a substitute for human judgement.\n",stderr);
ali@0	3695	fputs("\n",stderr);
ali@0	3696	}

author	ali <ali@juiblex.co.uk>
	Sun Oct 27 17:01:47 2013 +0000 (2013-10-27)
changeset 103	d22d8cd4f628
parent 102	ff0aa9b1397a
child 104	70cc629ec1e0
permissions	-rw-r--r--