bookloupe-testing: bookloupe/bookloupe.c@d14a95cb0aab (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@99	30	#include "bookloupe.h"
ali@99	31	#include "counters.h"
ali@103	32	#include "pending.h"
ali@71	33	#include "HTMLentities.h"
ali@0	34
ali@166	35	gchar charset; / Or NULL for auto (ISO_8859-1/ASCII or UNICODE) */
ali@166	36	GIConv charset_validator=(GIConv)-1;
ali@166	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@166	133	gchar *opt_charset;
ali@0	134
ali@167	135	gboolean typo_compat,paranoid_compat;
ali@167	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@167	140	{ "no-dp", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	141	G_OPTION_ARG_NONE, pswit+DP_SWITCH,
ali@167	142	"Don't ignore DP-specific markup", NULL },
ali@167	143	{ "echo", 0, G_OPTION_FLAG_HIDDEN, G_OPTION_ARG_NONE, pswit+ECHO_SWITCH,
ali@167	144	"Echo queried line", NULL },
ali@167	145	{ "no-echo", 'e', G_OPTION_FLAG_REVERSE,
ali@167	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@167	150	{ "no-squote", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	151	G_OPTION_ARG_NONE, pswit+SQUOTE_SWITCH,
ali@167	152	"Don't check single quotes", NULL },
ali@167	153	{ "typo", 0, 0, G_OPTION_ARG_NONE, pswit+TYPO_SWITCH,
ali@69	154	"Check common typos", NULL },
ali@167	155	{ "no-typo", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	156	G_OPTION_ARG_NONE, pswit+TYPO_SWITCH,
ali@167	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@167	160	{ "no-qpara", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	161	G_OPTION_ARG_NONE, pswit+QPARA_SWITCH,
ali@167	162	"Don't require closure of quotes on every paragraph", NULL },
ali@167	163	{ "paranoid", 0, G_OPTION_FLAG_HIDDEN,
ali@167	164	G_OPTION_ARG_NONE, pswit+PARANOID_SWITCH,
ali@167	165	"Enable paranoid querying of everything", NULL },
ali@167	166	{ "no-paranoid", 0, G_OPTION_FLAG_REVERSE,
ali@167	167	G_OPTION_ARG_NONE, pswit+PARANOID_SWITCH,
ali@69	168	"Disable paranoid querying of everything", NULL },
ali@167	169	{ "line-end", 0, G_OPTION_FLAG_HIDDEN,
ali@167	170	G_OPTION_ARG_NONE, pswit+LINE_END_SWITCH,
ali@167	171	"Enable line end checking", NULL },
ali@167	172	{ "no-line-end", 'l', G_OPTION_FLAG_REVERSE,
ali@167	173	G_OPTION_ARG_NONE, pswit+LINE_END_SWITCH,
ali@167	174	"Diable 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@167	177	{ "no-overview", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	178	G_OPTION_ARG_NONE, pswit+OVERVIEW_SWITCH,
ali@167	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@167	182	{ "no-stdout", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	183	G_OPTION_ARG_NONE, pswit+STDOUT_SWITCH,
ali@167	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@167	187	{ "no-header", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	188	G_OPTION_ARG_NONE, pswit+HEADER_SWITCH,
ali@167	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@167	192	{ "no-markup", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	193	G_OPTION_ARG_NONE, pswit+MARKUP_SWITCH,
ali@167	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@167	197	{ "no-usertypo", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	198	G_OPTION_ARG_NONE, pswit+USERTYPO_SWITCH,
ali@167	199	"Ignore file of user-defined typos", NULL },
ali@167	200	{ "verbose", 'v', 0, G_OPTION_ARG_NONE, pswit+VERBOSE_SWITCH,
ali@167	201	"Verbose - list everything", NULL },
ali@167	202	{ "no-verbose", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@167	203	G_OPTION_ARG_NONE, pswit+VERBOSE_SWITCH,
ali@167	204	"Switch off verbose mode", NULL },
ali@168	205	{ "charset", 0, 0, G_OPTION_ARG_STRING, &opt_charset,
ali@168	206	"Set of characters valid for this ebook", "NAME" },
ali@167	207	{ NULL }
ali@167	208	};
ali@167	209
ali@167	210	/*
ali@167	211	* Options relating to configuration which make no sense from inside
ali@167	212	* a configuration file.
ali@167	213	*/
ali@167	214
ali@167	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@167	218	{ "dump-config", 0, 0, G_OPTION_ARG_NONE, pswit+DUMP_CONFIG_SWITCH,
ali@167	219	"Dump current config settings", NULL },
ali@167	220	{ NULL }
ali@167	221	};
ali@167	222
ali@167	223	static GOptionEntry compatibility_options[]={
ali@167	224	{ "toggle-typo", 't', 0, G_OPTION_ARG_NONE, &typo_compat,
ali@167	225	"Toggle checking for common typos", NULL },
ali@167	226	{ "toggle-relaxed", 'x', 0, G_OPTION_ARG_NONE, &paranoid_compat,
ali@167	227	"Toggle both paranoid mode and common typos", NULL },
ali@69	228	{ NULL }
ali@69	229	};
ali@0	230
ali@142	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@69	254	char flgets(char *,long);
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@169	271	gboolean set_charset(const char name,GError *err)
ali@169	272	{
ali@169	273	/* The various UNICODE encodings all share the same character set. */
ali@169	274	const char *unicode_aliases[]={ "UCS-2", "UCS-2BE", "UCS-2LE", "UCS-4",
ali@169	275	"UCS-4BE", "UCS-4LE", "UCS2", "UCS4", "UNICODE", "UNICODEBIG",
ali@169	276	"UNICODELITTLE", "UTF-7", "UTF-8", "UTF-16", "UTF-16BE", "UTF-16LE",
ali@169	277	"UTF-32", "UTF-32BE", "UTF-32LE", "UTF7", "UTF8", "UTF16", "UTF16BE",
ali@169	278	"UTF16LE", "UTF32", "UTF32BE", "UTF32LE" };
ali@169	279	int i;
ali@169	280	if (charset)
ali@169	281	g_free(charset);
ali@169	282	if (charset_validator!=(GIConv)-1)
ali@169	283	g_iconv_close(charset_validator);
ali@169	284	if (!name \|\| !g_strcasecmp(name,"auto"))
ali@169	285	{
ali@169	286	charset=NULL;
ali@169	287	charset_validator=(GIConv)-1;
ali@169	288	return TRUE;
ali@169	289	}
ali@169	290	else
ali@169	291	charset=g_strdup(name);
ali@169	292	for(i=0;i<G_N_ELEMENTS(unicode_aliases);i++)
ali@169	293	if (!g_strcasecmp(charset,unicode_aliases[i]))
ali@169	294	{
ali@169	295	g_free(charset);
ali@169	296	charset=g_strdup("UTF-8");
ali@169	297	break;
ali@169	298	}
ali@169	299	if (!strcmp(charset,"UTF-8"))
ali@169	300	charset_validator=(GIConv)-1;
ali@169	301	else
ali@169	302	{
ali@169	303	charset_validator=g_iconv_open(charset,"UTF-8");
ali@169	304	if (charset_validator==(GIConv)-1)
ali@169	305	{
ali@169	306	g_set_error(err,G_CONVERT_ERROR,G_CONVERT_ERROR_NO_CONVERSION,
ali@169	307	"Unknown character set \"%s\"",charset);
ali@169	308	return FALSE;
ali@169	309	}
ali@169	310	}
ali@169	311	return TRUE;
ali@169	312	}
ali@169	313
ali@167	314	GKeyFile *config;
ali@167	315
ali@167	316	void config_file_update(GKeyFile *kf)
ali@167	317	{
ali@167	318	int i;
ali@169	319	const char *s;
ali@167	320	gboolean sw;
ali@167	321	for(i=0;options[i].long_name;i++)
ali@167	322	{
ali@167	323	if (g_str_has_prefix(options[i].long_name,"no-"))
ali@167	324	continue;
ali@167	325	if (options[i].arg==G_OPTION_ARG_NONE)
ali@167	326	{
ali@167	327	sw=(gboolean )options[i].arg_data;
ali@167	328	if (options[i].flags&G_OPTION_FLAG_REVERSE)
ali@167	329	sw=!sw;
ali@167	330	g_key_file_set_boolean(kf,"options",options[i].long_name,sw);
ali@167	331	}
ali@169	332	else if (options[i].arg==G_OPTION_ARG_STRING)
ali@169	333	{
ali@169	334	s=(gchar *)options[i].arg_data;
ali@169	335	if (!s)
ali@169	336	s="auto";
ali@169	337	g_key_file_set_string(kf,"options",options[i].long_name,s);
ali@169	338	}
ali@167	339	else
ali@167	340	g_assert_not_reached();
ali@167	341	}
ali@167	342	}
ali@167	343
ali@167	344	void config_file_add_comments(GKeyFile *kf)
ali@167	345	{
ali@167	346	int i;
ali@167	347	gchar *comment;
ali@167	348	g_key_file_set_comment(kf,NULL,NULL," Default configuration for bookloupe",
ali@167	349	NULL);
ali@167	350	for(i=0;options[i].long_name;i++)
ali@167	351	{
ali@167	352	if (g_str_has_prefix(options[i].long_name,"no-"))
ali@167	353	continue;
ali@167	354	comment=g_strconcat(" ",options[i].description,NULL);
ali@167	355	g_key_file_set_comment(kf,"options",options[i].long_name,comment,NULL);
ali@167	356	g_free(comment);
ali@167	357	}
ali@167	358	}
ali@167	359
ali@167	360	void dump_config(void)
ali@167	361	{
ali@167	362	gchar *s;
ali@167	363	if (config)
ali@167	364	config_file_update(config);
ali@167	365	else
ali@167	366	{
ali@167	367	config=g_key_file_new();
ali@167	368	config_file_update(config);
ali@167	369	config_file_add_comments(config);
ali@167	370	}
ali@167	371	s=g_key_file_to_data(config,NULL,NULL);
ali@167	372	if (s)
ali@167	373	g_print("%s",s);
ali@167	374	g_free(s);
ali@167	375	}
ali@167	376
ali@167	377	GKeyFile read_config_file(gchar *full_path)
ali@167	378	{
ali@167	379	int i;
ali@167	380	GError *err=NULL;
ali@167	381	gchar **search_dirs;
ali@167	382	gchar *path;
ali@167	383	const char *search_path;
ali@167	384	GKeyFile *kf;
ali@167	385	kf=g_key_file_new();
ali@167	386	search_path=g_getenv("BOOKLOUPE_CONFIG_PATH");
ali@167	387	if (search_path)
ali@167	388	{
ali@167	389	#ifdef __WIN32__
ali@167	390	search_dirs=g_strsplit(search_path,";",0);
ali@167	391	#else
ali@167	392	search_dirs=g_strsplit(search_path,":",0);
ali@167	393	#endif
ali@167	394	}
ali@167	395	else
ali@167	396	{
ali@167	397	search_dirs=g_new(gchar *,4);
ali@167	398	search_dirs[0]=g_get_current_dir();
ali@167	399	search_dirs[1]=g_strdup(running_from);
ali@167	400	search_dirs[2]=g_strdup(g_get_user_config_dir());
ali@167	401	search_dirs[3]=NULL;
ali@167	402	}
ali@167	403	for(i=0;search_dirs[i];i++)
ali@167	404	{
ali@167	405	path=g_build_filename(search_dirs[i],"bookloupe.ini",NULL);
ali@167	406	if (g_key_file_load_from_file(kf,path,
ali@167	407	G_KEY_FILE_KEEP_COMMENTS\|G_KEY_FILE_KEEP_TRANSLATIONS,&err))
ali@167	408	break;
ali@167	409	if (!g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@167	410	{
ali@167	411	g_printerr("Bookloupe: Error reading %s\n",path);
ali@167	412	g_printerr("%s\n",err->message);
ali@167	413	exit(1);
ali@167	414	}
ali@167	415	g_clear_error(&err);
ali@167	416	g_free(path);
ali@167	417	path=NULL;
ali@167	418	}
ali@167	419	if (!search_dirs[i])
ali@167	420	{
ali@167	421	g_key_file_free(kf);
ali@167	422	kf=NULL;
ali@167	423	}
ali@167	424	g_strfreev(search_dirs);
ali@167	425	if (full_path && kf)
ali@167	426	*full_path=path;
ali@167	427	else
ali@167	428	g_free(path);
ali@167	429	return kf;
ali@167	430	}
ali@167	431
ali@167	432	void parse_config_file(void)
ali@167	433	{
ali@167	434	int i,j;
ali@169	435	gchar path,s;
ali@167	436	gchar **keys;
ali@167	437	gboolean sw;
ali@167	438	GError *err=NULL;
ali@167	439	config=read_config_file(&path);
ali@167	440	if (config)
ali@167	441	keys=g_key_file_get_keys(config,"options",NULL,NULL);
ali@167	442	else
ali@167	443	keys=NULL;
ali@167	444	if (keys)
ali@167	445	{
ali@167	446	for(i=0;keys[i];i++)
ali@167	447	{
ali@167	448	for(j=0;options[j].long_name;j++)
ali@167	449	{
ali@167	450	if (g_str_has_prefix(options[j].long_name,"no-"))
ali@167	451	continue;
ali@167	452	else if (!strcmp(keys[i],options[j].long_name))
ali@167	453	{
ali@167	454	if (options[j].arg==G_OPTION_ARG_NONE)
ali@167	455	{
ali@167	456	sw=g_key_file_get_boolean(config,"options",keys[i],
ali@167	457	&err);
ali@167	458	if (err)
ali@167	459	{
ali@167	460	g_printerr("Bookloupe: %s: options.%s: %s\n",
ali@167	461	path,keys[i],err->message);
ali@167	462	g_clear_error(&err);
ali@167	463	}
ali@169	464	else
ali@169	465	{
ali@169	466	if (options[j].flags&G_OPTION_FLAG_REVERSE)
ali@169	467	sw=!sw;
ali@169	468	(gboolean )options[j].arg_data=sw;
ali@169	469	}
ali@169	470	break;
ali@169	471	}
ali@169	472	else if (options[j].arg==G_OPTION_ARG_STRING)
ali@169	473	{
ali@169	474	s=g_key_file_get_string(config,"options",keys[i],
ali@169	475	&err);
ali@169	476	if (err)
ali@169	477	{
ali@169	478	g_printerr("Bookloupe: %s: options.%s: %s\n",
ali@169	479	path,keys[i],err->message);
ali@169	480	g_clear_error(&err);
ali@169	481	}
ali@169	482	else
ali@169	483	{
ali@169	484	g_free((gchar *)options[j].arg_data);
ali@169	485	if (!g_strcmp0(s,"auto"))
ali@169	486	{
ali@169	487	(gchar *)options[j].arg_data=NULL;
ali@169	488	g_free(s);
ali@169	489	}
ali@169	490	else
ali@169	491	(gchar *)options[j].arg_data=s;
ali@169	492	}
ali@167	493	break;
ali@167	494	}
ali@167	495	else
ali@167	496	g_assert_not_reached();
ali@167	497	}
ali@167	498	}
ali@167	499	if (!options[j].long_name)
ali@167	500	g_printerr("Bookloupe: %s: Unknown option \"%s\" ignored\n",
ali@167	501	path,keys[i]);
ali@167	502	}
ali@167	503	g_strfreev(keys);
ali@167	504	}
ali@167	505	if (config)
ali@167	506	g_free(path);
ali@167	507	}
ali@167	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@167	513	GOptionGroup *compatibility;
ali@69	514	context=g_option_context_new(
ali@167	515	"file - look for errors in Project Gutenberg(TM) etexts");
ali@69	516	g_option_context_add_main_entries(context,options,NULL);
ali@167	517	g_option_context_add_main_entries(context,config_options,NULL);
ali@167	518	compatibility=g_option_group_new("compatibility",
ali@167	519	"Options for Compatibility with Gutcheck:",
ali@167	520	"Show compatibility options",NULL,NULL);
ali@167	521	g_option_group_add_entries(compatibility,compatibility_options);
ali@167	522	g_option_context_add_group(context,compatibility);
ali@167	523	g_option_context_set_description(context,
ali@167	524	"For simplicity, only the switch options which reverse the\n"
ali@167	525	"default configuration are listed. In most cases, both vanilla\n"
ali@167	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@167	533	if (typo_compat)
ali@69	534	pswit[TYPO_SWITCH]=!pswit[TYPO_SWITCH];
ali@167	535	if (paranoid_compat)
ali@167	536	{
ali@167	537	pswit[PARANOID_SWITCH]=!pswit[PARANOID_SWITCH];
ali@167	538	pswit[TYPO_SWITCH]=!pswit[TYPO_SWITCH];
ali@167	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@166	561	if (opt_charset && !set_charset(opt_charset,&err))
ali@166	562	{
ali@166	563	g_printerr("%s\n",err->message);
ali@166	564	exit(1);
ali@166	565	}
ali@167	566	if (pswit[DUMP_CONFIG_SWITCH])
ali@167	567	{
ali@167	568	dump_config();
ali@167	569	exit(0);
ali@167	570	}
ali@166	571	g_free(opt_charset);
ali@166	572	opt_charset=NULL;
ali@167	573	if (pswit[OVERVIEW_SWITCH])
ali@167	574	/* just print summary; don't echo */
ali@167	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@166	635	{
ali@72	636	utf8=g_utf8_normalize(contents,len,G_NORMALIZE_DEFAULT_COMPOSE);
ali@166	637	if (!charset)
ali@166	638	(void)set_charset("UNICODE",NULL);
ali@166	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@167	722	/* Paranoid checking is turned OFF, not on, by its switch */
ali@167	723	pswit[PARANOID_SWITCH]=TRUE;
ali@167	724	/* if running in paranoid mode, typo checks default to enabled */
ali@167	725	pswit[TYPO_SWITCH]=TRUE;
ali@167	726	/* Line-end checking is turned OFF, not on, by its switch */
ali@167	727	pswit[LINE_END_SWITCH]=TRUE;
ali@167	728	/* Echoing is turned OFF, not on, by its switch */
ali@167	729	pswit[ECHO_SWITCH]=TRUE;
ali@167	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@142	749	if (cnt_quote)
ali@142	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@142	765	cnt_quote+cnt_brack+cnt_bin+cnt_odd+cnt_long+cnt_short+cnt_punct+
ali@142	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@166	771	set_charset(NULL,NULL);
ali@167	772	if (config)
ali@167	773	g_key_file_free(config);
ali@40	774	return 0;
ali@0	775	}
ali@0	776
ali@147	777	void count_dashes(const char line,const char dash,
ali@147	778	struct dash_results *results)
ali@147	779	{
ali@147	780	int i;
ali@147	781	gchar **tokens;
ali@147	782	gunichar pc,nc;
ali@147	783	gboolean spaced=FALSE,unspaced=FALSE,spaced2=FALSE;
ali@147	784	if (!*line)
ali@147	785	return;
ali@147	786	tokens=g_strsplit(line,dash,0);
ali@147	787	if (tokens[1])
ali@147	788	results->base++;
ali@147	789	for(i=1;tokens[i];i++)
ali@147	790	{
ali@147	791	pc=g_utf8_get_char(g_utf8_prev_char(tokens[i-1]+strlen(tokens[i-1])));
ali@147	792	nc=g_utf8_get_char(tokens[i]);
ali@147	793	if (g_unichar_isspace(pc) \|\| g_unichar_isspace(nc))
ali@147	794	spaced=TRUE;
ali@147	795	if (g_unichar_isspace(pc) && g_unichar_isspace(nc))
ali@147	796	spaced2=TRUE;
ali@147	797	else if (!g_unichar_isspace(pc) && !g_unichar_isspace(nc))
ali@147	798	unspaced=TRUE;
ali@147	799	}
ali@147	800	if (spaced)
ali@147	801	results->space++;
ali@147	802	if (spaced2)
ali@147	803	/* count of lines with em-dashes with spaces both sides */
ali@147	804	results->non_PG_space++;
ali@147	805	if (unspaced)
ali@147	806	/* count of lines with PG-type em-dashes with no spaces */
ali@147	807	results->PG_space++;
ali@147	808	g_strfreev(tokens);
ali@147	809	}
ali@147	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@147	829	struct dash_results tmp_dash_results;
ali@69	830	gchar *inword;
ali@142	831	QuoteClass qc;
ali@69	832	lines=g_strsplit(etext,"\n",0);
ali@69	833	for (j=0;lines[j];j++)
ali@40	834	{
ali@70	835	lbytes=strlen(lines[j]);
ali@82	836	while (lbytes>0 && lines[j][lbytes-1]=='\r')
ali@70	837	lines[j][--lbytes]='\0';
ali@70	838	llen=g_utf8_strlen(lines[j],lbytes);
ali@68	839	linecnt++;
ali@69	840	if (strstr(lines[j],"*END") && strstr(lines[j],"SMALL PRINT") &&
ali@69	841	(strstr(lines[j],"PUBLIC DOMAIN") \|\| strstr(lines[j],"COPYRIGHT")))
ali@40	842	{
ali@68	843	if (spline)
ali@70	844	g_print(" --> Duplicate header?\n");
ali@68	845	spline=linecnt+1; /* first line of non-header text, that is */
ali@40	846	}
ali@69	847	if (!strncmp(lines[j],"*** START",9) &&
ali@69	848	strstr(lines[j],"PROJECT GUTENBERG"))
ali@40	849	{
ali@68	850	if (nspline)
ali@70	851	g_print(" --> Duplicate header?\n");
ali@68	852	nspline=linecnt+1; /* first line of non-header text, that is */
ali@40	853	}
ali@68	854	if (spline \|\| nspline)
ali@40	855	{
ali@70	856	lc_line=g_utf8_strdown(lines[j],lbytes);
ali@69	857	if (strstr(lc_line,"end") && strstr(lc_line,"project gutenberg"))
ali@40	858	{
ali@69	859	if (strstr(lc_line,"end")<strstr(lc_line,"project gutenberg"))
ali@40	860	{
ali@68	861	if (results.footerline)
ali@40	862	{
ali@40	863	/* it's an old-form header - we can detect duplicates */
ali@68	864	if (!nspline)
ali@70	865	g_print(" --> Duplicate footer?\n");
ali@40	866	}
ali@68	867	else
ali@68	868	results.footerline=linecnt;
ali@40	869	}
ali@40	870	}
ali@69	871	g_free(lc_line);
ali@40	872	}
ali@68	873	if (spline)
ali@41	874	results.firstline=spline;
ali@68	875	if (nspline)
ali@41	876	results.firstline=nspline; /* override with new */
ali@68	877	if (results.footerline)
ali@40	878	continue; /* don't count the boilerplate in the footer */
ali@68	879	results.totlen+=llen;
ali@70	880	for (s=lines[j];*s;s=g_utf8_next_char(s))
ali@40	881	{
ali@70	882	if (g_utf8_get_char(s)>127)
ali@41	883	results.binlen++;
ali@70	884	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@41	885	results.alphalen++;
ali@142	886	if (s>lines[j])
ali@142	887	{
ali@142	888	if (CHAR_IS_DQUOTE(g_utf8_get_char(s)))
ali@142	889	qc=QUOTE_CLASS(g_utf8_get_char(s));
ali@142	890	else
ali@142	891	qc=INVALID_QUOTE;
ali@142	892	if ((qc==CLOSING_QUOTE \|\| qc==NEUTRAL_QUOTE) &&
ali@147	893	g_unichar_isalpha(g_utf8_get_char(g_utf8_prev_char(s))))
ali@142	894	results.endquote_count++;
ali@142	895	}
ali@40	896	}
ali@69	897	if (llen>2 && lastlen>2 && lastlen<SHORTEST_PG_LINE && lastblen>2 &&
ali@69	898	lastblen>SHORTEST_PG_LINE && laststart!=CHAR_SPACE)
ali@41	899	results.shortline++;
ali@70	900	if (lbytes>0 &&
ali@70	901	g_utf8_get_char(g_utf8_prev_char(lines[j]+lbytes))<=CHAR_SPACE)
ali@40	902	cnt_spacend++;
ali@69	903	if (strstr(lines[j],".,"))
ali@41	904	results.dotcomma++;
ali@68	905	/* only count ast lines for ignoring purposes where there is */
ali@68	906	/* locase text on the line */
ali@69	907	if (strchr(lines[j],'*'))
ali@40	908	{
ali@70	909	for (s=lines[j];*s;s=g_utf8_next_char(s))
ali@70	910	if (g_unichar_islower(g_utf8_get_char(s)))
ali@68	911	break;
ali@70	912	if (*s)
ali@41	913	results.astline++;
ali@40	914	}
ali@69	915	if (strchr(lines[j],'/'))
ali@68	916	results.fslashline++;
ali@82	917	if (lbytes>0)
ali@82	918	{
ali@82	919	for (s=g_utf8_prev_char(lines[j]+lbytes);
ali@82	920	s>lines[j] && g_utf8_get_char(s)<=CHAR_SPACE;
ali@82	921	s=g_utf8_prev_char(s))
ali@82	922	;
ali@82	923	if (s>g_utf8_next_char(lines[j]) && g_utf8_get_char(s)=='-' &&
ali@82	924	g_utf8_get_char(g_utf8_prev_char(s))!='-')
ali@82	925	results.hyphens++;
ali@82	926	}
ali@68	927	if (llen>LONGEST_PG_LINE)
ali@41	928	results.longline++;
ali@68	929	if (llen>WAY_TOO_LONG)
ali@41	930	results.verylongline++;
ali@69	931	if (strchr(lines[j],'<') && strchr(lines[j],'>'))
ali@40	932	{
ali@69	933	i=(int)(strchr(lines[j],'>')-strchr(lines[j],'<')+1);
ali@68	934	if (i>0)
ali@68	935	results.htmcount++;
ali@69	936	if (strstr(lines[j],"<i>"))
ali@41	937	results.htmcount+=4; /* bonus marks! */
ali@40	938	}
ali@68	939	/* Check for spaced em-dashes */
ali@147	940	memset(&tmp_dash_results,0,sizeof(tmp_dash_results));
ali@147	941	count_dashes(lines[j],"--",&tmp_dash_results);
ali@147	942	count_dashes(lines[j],"—",&tmp_dash_results);
ali@147	943	if (tmp_dash_results.base)
ali@147	944	results.emdash.base++;
ali@147	945	if (tmp_dash_results.non_PG_space)
ali@147	946	results.emdash.non_PG_space++;
ali@147	947	if (tmp_dash_results.PG_space)
ali@147	948	results.emdash.PG_space++;
ali@69	949	for (s=lines[j];*s;)
ali@40	950	{
ali@69	951	inword=getaword(&s);
ali@68	952	if (!strcmp(inword,"hij") \|\| !strcmp(inword,"niet"))
ali@68	953	results.Dutchcount++;
ali@68	954	if (!strcmp(inword,"dans") \|\| !strcmp(inword,"avec"))
ali@68	955	results.Frenchcount++;
ali@68	956	if (!strcmp(inword,"0") \|\| !strcmp(inword,"1"))
ali@68	957	results.standalone_digit++;
ali@69	958	g_free(inword);
ali@40	959	}
ali@68	960	/* Check for spaced dashes */
ali@69	961	if (strstr(lines[j]," -") && *(strstr(lines[j]," -")+2)!='-')
ali@41	962	results.spacedash++;
ali@68	963	lastblen=lastlen;
ali@69	964	lastlen=llen;
ali@69	965	laststart=lines[j][0];
ali@40	966	}
ali@69	967	g_strfreev(lines);
ali@41	968	return &results;
ali@41	969	}
ali@41	970
ali@42	971	/*
ali@42	972	* report_first_pass:
ali@42	973	*
ali@42	974	* Make some snap decisions based on the first pass results.
ali@42	975	*/
ali@42	976	struct warnings report_first_pass(struct first_pass_results results)
ali@42	977	{
ali@42	978	static struct warnings warnings={0};
ali@42	979	if (cnt_spacend>0)
ali@70	980	g_print(" --> %ld lines in this file have white space at end\n",
ali@42	981	cnt_spacend);
ali@42	982	warnings.dotcomma=1;
ali@42	983	if (results->dotcomma>5)
ali@42	984	{
ali@68	985	warnings.dotcomma=0;
ali@70	986	g_print(" --> %ld lines in this file contain '.,'. "
ali@42	987	"Not reporting them.\n",results->dotcomma);
ali@42	988	}
ali@42	989	/*
ali@42	990	* If more than 50 lines, or one-tenth, are short,
ali@42	991	* don't bother reporting them.
ali@42	992	*/
ali@42	993	warnings.shortline=1;
ali@42	994	if (results->shortline>50 \|\| results->shortline*10>linecnt)
ali@42	995	{
ali@68	996	warnings.shortline=0;
ali@70	997	g_print(" --> %ld lines in this file are short. "
ali@42	998	"Not reporting short lines.\n",results->shortline);
ali@42	999	}
ali@42	1000	/*
ali@42	1001	* If more than 50 lines, or one-tenth, are long,
ali@42	1002	* don't bother reporting them.
ali@42	1003	*/
ali@42	1004	warnings.longline=1;
ali@42	1005	if (results->longline>50 \|\| results->longline*10>linecnt)
ali@42	1006	{
ali@68	1007	warnings.longline=0;
ali@70	1008	g_print(" --> %ld lines in this file are long. "
ali@42	1009	"Not reporting long lines.\n",results->longline);
ali@42	1010	}
ali@42	1011	/* If more than 10 lines contain asterisks, don't bother reporting them. */
ali@42	1012	warnings.ast=1;
ali@42	1013	if (results->astline>10)
ali@42	1014	{
ali@68	1015	warnings.ast=0;
ali@70	1016	g_print(" --> %ld lines in this file contain asterisks. "
ali@42	1017	"Not reporting them.\n",results->astline);
ali@42	1018	}
ali@42	1019	/*
ali@42	1020	* If more than 10 lines contain forward slashes,
ali@42	1021	* don't bother reporting them.
ali@42	1022	*/
ali@42	1023	warnings.fslash=1;
ali@42	1024	if (results->fslashline>10)
ali@42	1025	{
ali@68	1026	warnings.fslash=0;
ali@70	1027	g_print(" --> %ld lines in this file contain forward slashes. "
ali@42	1028	"Not reporting them.\n",results->fslashline);
ali@42	1029	}
ali@42	1030	/*
ali@42	1031	* If more than 20 lines contain unpunctuated endquotes,
ali@42	1032	* don't bother reporting them.
ali@42	1033	*/
ali@42	1034	warnings.endquote=1;
ali@42	1035	if (results->endquote_count>20)
ali@42	1036	{
ali@68	1037	warnings.endquote=0;
ali@70	1038	g_print(" --> %ld lines in this file contain unpunctuated endquotes. "
ali@42	1039	"Not reporting them.\n",results->endquote_count);
ali@42	1040	}
ali@42	1041	/*
ali@42	1042	* If more than 15 lines contain standalone digits,
ali@42	1043	* don't bother reporting them.
ali@42	1044	*/
ali@42	1045	warnings.digit=1;
ali@42	1046	if (results->standalone_digit>10)
ali@42	1047	{
ali@68	1048	warnings.digit=0;
ali@70	1049	g_print(" --> %ld lines in this file contain standalone 0s and 1s. "
ali@42	1050	"Not reporting them.\n",results->standalone_digit);
ali@42	1051	}
ali@42	1052	/*
ali@42	1053	* If more than 20 lines contain hyphens at end,
ali@42	1054	* don't bother reporting them.
ali@42	1055	*/
ali@42	1056	warnings.hyphen=1;
ali@42	1057	if (results->hyphens>20)
ali@42	1058	{
ali@68	1059	warnings.hyphen=0;
ali@70	1060	g_print(" --> %ld lines in this file have hyphens at end. "
ali@42	1061	"Not reporting them.\n",results->hyphens);
ali@42	1062	}
ali@42	1063	if (results->htmcount>20 && !pswit[MARKUP_SWITCH])
ali@42	1064	{
ali@70	1065	g_print(" --> Looks like this is HTML. Switching HTML mode ON.\n");
ali@68	1066	pswit[MARKUP_SWITCH]=1;
ali@42	1067	}
ali@42	1068	if (results->verylongline>0)
ali@70	1069	g_print(" --> %ld lines in this file are VERY long!\n",
ali@42	1070	results->verylongline);
ali@42	1071	/*
ali@42	1072	* If there are more non-PG spaced dashes than PG em-dashes,
ali@42	1073	* assume it's deliberate.
ali@42	1074	* Current PG guidelines say don't use them, but older texts do,
ali@42	1075	* and some people insist on them whatever the guidelines say.
ali@42	1076	*/
ali@42	1077	warnings.dash=1;
ali@147	1078	if (results->spacedash+results->emdash.non_PG_space>
ali@147	1079	results->emdash.PG_space)
ali@42	1080	{
ali@68	1081	warnings.dash=0;
ali@70	1082	g_print(" --> There are %ld spaced dashes and em-dashes. "
ali@42	1083	"Not reporting them.\n",
ali@147	1084	results->spacedash+results->emdash.non_PG_space);
ali@42	1085	}
ali@166	1086	if (charset)
ali@166	1087	warnings.bin=0;
ali@166	1088	else
ali@42	1089	{
ali@166	1090	/* Charset ISO_8859-1/ASCII checks for compatibility with gutcheck */
ali@166	1091	warnings.bin=1;
ali@166	1092	/* If more than a quarter of characters are hi-bit, bug out. */
ali@166	1093	if (results->binlen*4>results->totlen)
ali@166	1094	{
ali@166	1095	g_print(" --> This file does not appear to be ASCII. "
ali@166	1096	"Terminating. Best of luck with it!\n");
ali@166	1097	exit(1);
ali@166	1098	}
ali@166	1099	if (results->alphalen*4<results->totlen)
ali@166	1100	{
ali@166	1101	g_print(" --> This file does not appear to be text. "
ali@166	1102	"Terminating. Best of luck with it!\n");
ali@166	1103	exit(1);
ali@166	1104	}
ali@166	1105	if (results->binlen*100>results->totlen \|\| results->binlen>100)
ali@166	1106	{
ali@166	1107	g_print(" --> There are a lot of foreign letters here. "
ali@166	1108	"Not reporting them.\n");
ali@166	1109	if (!pswit[VERBOSE_SWITCH])
ali@166	1110	warnings.bin=0;
ali@166	1111	}
ali@42	1112	}
ali@69	1113	warnings.isDutch=FALSE;
ali@42	1114	if (results->Dutchcount>50)
ali@42	1115	{
ali@69	1116	warnings.isDutch=TRUE;
ali@70	1117	g_print(" --> This looks like Dutch - "
ali@42	1118	"switching off dashes and warnings for 's Middags case.\n");
ali@42	1119	}
ali@69	1120	warnings.isFrench=FALSE;
ali@42	1121	if (results->Frenchcount>50)
ali@42	1122	{
ali@69	1123	warnings.isFrench=TRUE;
ali@70	1124	g_print(" --> This looks like French - "
ali@42	1125	"switching off some doublepunct.\n");
ali@42	1126	}
ali@42	1127	if (results->firstline && results->footerline)
ali@70	1128	g_print(" The PG header and footer appear to be already on.\n");
ali@42	1129	else
ali@42	1130	{
ali@68	1131	if (results->firstline)
ali@70	1132	g_print(" The PG header is on - no footer.\n");
ali@68	1133	if (results->footerline)
ali@70	1134	g_print(" The PG footer is on - no header.\n");
ali@42	1135	}
ali@70	1136	g_print("\n");
ali@42	1137	if (pswit[VERBOSE_SWITCH])
ali@42	1138	{
ali@68	1139	warnings.shortline=1;
ali@68	1140	warnings.dotcomma=1;
ali@68	1141	warnings.longline=1;
ali@68	1142	warnings.dash=1;
ali@68	1143	warnings.digit=1;
ali@68	1144	warnings.ast=1;
ali@68	1145	warnings.fslash=1;
ali@68	1146	warnings.hyphen=1;
ali@68	1147	warnings.endquote=1;
ali@70	1148	g_print(" * Verbose output is ON -- you asked for it! *\n");
ali@42	1149	}
ali@42	1150	if (warnings.isDutch)
ali@68	1151	warnings.dash=0;
ali@42	1152	if (results->footerline>0 && results->firstline>0 &&
ali@42	1153	results->footerline>results->firstline &&
ali@42	1154	results->footerline-results->firstline<100)
ali@42	1155	{
ali@70	1156	g_print(" --> I don't really know where this text starts. \n");
ali@70	1157	g_print(" There are no reference points.\n");
ali@70	1158	g_print(" I'm going to have to report the header and footer "
ali@42	1159	"as well.\n");
ali@68	1160	results->firstline=0;
ali@42	1161	}
ali@42	1162	return &warnings;
ali@42	1163	}
ali@42	1164
ali@43	1165	/*
ali@43	1166	* analyse_quotes:
ali@43	1167	*
ali@43	1168	* Look along the line, accumulate the count of quotes, and see
ali@43	1169	* if this is an empty line - i.e. a line with nothing on it
ali@43	1170	* but spaces.
ali@43	1171	* If line has just spaces, period, * and/or - on it, don't
ali@43	1172	* count it, since empty lines with asterisks or dashes to
ali@43	1173	* separate sections are common.
ali@43	1174	*
ali@69	1175	* Returns: TRUE if the line is empty.
ali@43	1176	*/
ali@164	1177	gboolean analyse_quotes(const char aline,struct counters counters)
ali@43	1178	{
ali@68	1179	int guessquote=0;
ali@69	1180	/* assume the line is empty until proven otherwise */
ali@69	1181	gboolean isemptyline=TRUE;
ali@70	1182	const char s=aline,sprev,*snext;
ali@70	1183	gunichar c;
ali@70	1184	sprev=NULL;
ali@142	1185	GError *tmp_err=NULL;
ali@43	1186	while (*s)
ali@43	1187	{
ali@70	1188	snext=g_utf8_next_char(s);
ali@70	1189	c=g_utf8_get_char(s);
ali@142	1190	if (CHAR_IS_DQUOTE(c))
ali@142	1191	(void)count_quote(counters,c,QUOTE_CLASS(c),&tmp_err);
ali@142	1192	else if (CHAR_IS_SQUOTE(c) && pswit[SQUOTE_SWITCH])
ali@43	1193	{
ali@43	1194	if (s==aline)
ali@43	1195	{
ali@43	1196	/*
ali@142	1197	* At start of line, it can only be a quotation mark.
ali@43	1198	* Hardcode a very common exception!
ali@43	1199	*/
ali@70	1200	if (!g_str_has_prefix(snext,"tis") &&
ali@70	1201	!g_str_has_prefix(snext,"Tis"))
ali@142	1202	(void)count_quote(counters,c,NEUTRAL_QUOTE,&tmp_err);
ali@43	1203	}
ali@70	1204	else if (g_unichar_isalpha(g_utf8_get_char(sprev)) &&
ali@70	1205	g_unichar_isalpha(g_utf8_get_char(snext)))
ali@43	1206	/* Do nothing! it's definitely an apostrophe, not a quote */
ali@43	1207	;
ali@43	1208	/* it's outside a word - let's check it out */
ali@99	1209	else if (c==CHAR_OPEN_SQUOTE \|\| c==CHAR_LS_QUOTE \|\|
ali@70	1210	g_unichar_isalpha(g_utf8_get_char(snext)))
ali@43	1211	{
ali@142	1212	/* certainly looks like a quotation mark */
ali@70	1213	if (!g_str_has_prefix(snext,"tis") &&
ali@70	1214	!g_str_has_prefix(snext,"Tis"))
ali@43	1215	/* hardcode a very common exception! */
ali@142	1216	{
ali@142	1217	if (strchr(".?!,;:",g_utf8_get_char(sprev)))
ali@142	1218	(void)count_quote(counters,c,NEUTRAL_QUOTE,&tmp_err);
ali@142	1219	else
ali@142	1220	(void)count_quote(counters,c,OPENING_QUOTE,&tmp_err);
ali@142	1221	}
ali@43	1222	}
ali@43	1223	else
ali@43	1224	{
ali@142	1225	/* now - is it a quotation mark? */
ali@43	1226	guessquote=0; /* accumulate clues */
ali@70	1227	if (g_unichar_isalpha(g_utf8_get_char(sprev)))
ali@43	1228	{
ali@43	1229	/* it follows a letter - could be either */
ali@43	1230	guessquote++;
ali@70	1231	if (g_utf8_get_char(sprev)=='s')
ali@43	1232	{
ali@43	1233	/* looks like a plural apostrophe */
ali@43	1234	guessquote-=3;
ali@70	1235	if (g_utf8_get_char(snext)==CHAR_SPACE)
ali@70	1236	/* bonus marks! */
ali@43	1237	guessquote-=2;
ali@43	1238	}
ali@142	1239	if (innermost_quote_matches(counters,c))
ali@142	1240	/*
ali@142	1241	* Give it the benefit of some doubt,
ali@142	1242	* if a squote is already open.
ali@142	1243	*/
ali@142	1244	guessquote++;
ali@142	1245	else
ali@142	1246	guessquote--;
ali@142	1247	if (guessquote>=0)
ali@142	1248	(void)count_quote(counters,c,CLOSING_QUOTE,&tmp_err);
ali@43	1249	}
ali@43	1250	else
ali@142	1251	/* no adjacent letter - it must be a quote of some kind */
ali@142	1252	(void)count_quote(counters,c,NEUTRAL_QUOTE,&tmp_err);
ali@43	1253	}
ali@43	1254	}
ali@142	1255	if (tmp_err)
ali@142	1256	{
ali@142	1257	if (pswit[ECHO_SWITCH])
ali@142	1258	g_print("\n%s\n",aline);
ali@142	1259	if (!pswit[OVERVIEW_SWITCH])
ali@142	1260	g_print(" Line %ld column %ld - %s\n",
ali@142	1261	linecnt,g_utf8_pointer_to_offset(aline,s)+1,tmp_err->message);
ali@142	1262	g_clear_error(&tmp_err);
ali@142	1263	}
ali@70	1264	if (c!=CHAR_SPACE && c!='-' && c!='.' && c!=CHAR_ASTERISK &&
ali@70	1265	c!='\r' && c!='\n')
ali@69	1266	isemptyline=FALSE; /* ignore lines like * * * as spacers */
ali@70	1267	if (c==CHAR_UNDERSCORE)
ali@43	1268	counters->c_unders++;
ali@103	1269	if (c==CHAR_OPEN_SBRACK)
ali@103	1270	{
ali@103	1271	if (!matching_difference(counters,COUNTER_ILLUSTRATION) &&
ali@103	1272	!matching_difference(counters,c) && s==aline &&
ali@103	1273	g_str_has_prefix(s,"[Illustration:"))
ali@103	1274	increment_matching(counters,COUNTER_ILLUSTRATION,TRUE);
ali@103	1275	else
ali@103	1276	increment_matching(counters,c,TRUE);
ali@103	1277	}
ali@103	1278	else if (c==CHAR_OPEN_CBRACK \|\| c==CHAR_OPEN_RBRACK)
ali@99	1279	increment_matching(counters,c,TRUE);
ali@103	1280	if (c==CHAR_CLOSE_SBRACK)
ali@103	1281	{
ali@103	1282	if (!matching_count(counters,COUNTER_ILLUSTRATION,FALSE) &&
ali@103	1283	!matching_difference(counters,c) && !*snext)
ali@103	1284	increment_matching(counters,COUNTER_ILLUSTRATION,FALSE);
ali@103	1285	else
ali@103	1286	increment_matching(counters,c,FALSE);
ali@103	1287	}
ali@103	1288	else if (c==CHAR_CLOSE_CBRACK \|\| c==CHAR_CLOSE_RBRACK)
ali@99	1289	increment_matching(counters,c,FALSE);
ali@70	1290	sprev=s;
ali@70	1291	s=snext;
ali@43	1292	}
ali@43	1293	return isemptyline;
ali@43	1294	}
ali@43	1295
ali@41	1296	/*
ali@67	1297	* check_for_control_characters:
ali@67	1298	*
ali@67	1299	* Check for invalid or questionable characters in the line
ali@67	1300	* Anything above 127 is invalid for plain ASCII, and
ali@67	1301	* non-printable control characters should also be flagged.
ali@67	1302	* Tabs should generally not be there.
ali@67	1303	*/
ali@67	1304	void check_for_control_characters(const char *aline)
ali@67	1305	{
ali@70	1306	gunichar c;
ali@67	1307	const char *s;
ali@70	1308	for (s=aline;*s;s=g_utf8_next_char(s))
ali@67	1309	{
ali@70	1310	c=g_utf8_get_char(s);
ali@67	1311	if (c<CHAR_SPACE && c!=CHAR_LF && c!=CHAR_CR && c!=CHAR_TAB)
ali@67	1312	{
ali@67	1313	if (pswit[ECHO_SWITCH])
ali@70	1314	g_print("\n%s\n",aline);
ali@67	1315	if (!pswit[OVERVIEW_SWITCH])
ali@70	1316	g_print(" Line %ld column %ld - Control character %u\n",
ali@70	1317	linecnt,g_utf8_pointer_to_offset(s,aline)+1,c);
ali@67	1318	else
ali@67	1319	cnt_bin++;
ali@67	1320	}
ali@67	1321	}
ali@67	1322	}
ali@67	1323
ali@67	1324	/*
ali@44	1325	* check_for_odd_characters:
ali@44	1326	*
ali@44	1327	* Check for binary and other odd characters.
ali@44	1328	*/
ali@44	1329	void check_for_odd_characters(const char aline,const struct warnings warnings,
ali@69	1330	gboolean isemptyline)
ali@44	1331	{
ali@44	1332	/* Don't repeat multiple warnings on one line. */
ali@166	1333	gboolean eInvalidChar=FALSE,eTab=FALSE,eTilde=FALSE;
ali@70	1334	gboolean eCarat=FALSE,eFSlash=FALSE,eAst=FALSE;
ali@44	1335	const char *s;
ali@70	1336	gunichar c;
ali@166	1337	gsize nb;
ali@166	1338	gchar *t;
ali@70	1339	for (s=aline;*s;s=g_utf8_next_char(s))
ali@44	1340	{
ali@70	1341	c=g_utf8_get_char(s);
ali@166	1342	if (warnings->bin && !eInvalidChar &&
ali@166	1343	(c<CHAR_SPACE && c!='\t' && c!='\n' \|\| c>127))
ali@44	1344	{
ali@44	1345	if (pswit[ECHO_SWITCH])
ali@70	1346	g_print("\n%s\n",aline);
ali@44	1347	if (!pswit[OVERVIEW_SWITCH])
ali@70	1348	if (c>127 && c<160 \|\| c>255)
ali@70	1349	g_print(" Line %ld column %ld - "
ali@70	1350	"Non-ISO-8859 character %u\n",
ali@70	1351	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@44	1352	else
ali@70	1353	g_print(" Line %ld column %ld - "
ali@70	1354	"Non-ASCII character %u\n",
ali@70	1355	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@44	1356	else
ali@44	1357	cnt_bin++;
ali@166	1358	eInvalidChar=TRUE;
ali@166	1359	}
ali@166	1360	if (!eInvalidChar && charset)
ali@166	1361	{
ali@166	1362	if (charset_validator==(GIConv)-1)
ali@166	1363	{
ali@166	1364	if (!g_unichar_isdefined(c))
ali@166	1365	{
ali@166	1366	if (pswit[ECHO_SWITCH])
ali@166	1367	g_print("\n%s\n",aline);
ali@166	1368	if (!pswit[OVERVIEW_SWITCH])
ali@166	1369	g_print(" Line %ld column %ld - Unassigned UNICODE "
ali@166	1370	"code point U+%04" G_GINT32_MODIFIER "X\n",
ali@166	1371	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@166	1372	else
ali@166	1373	cnt_bin++;
ali@166	1374	eInvalidChar=TRUE;
ali@166	1375	}
ali@166	1376	else if (c>=0xE000 && c<=0xF8FF \|\| c>=0xF0000 && c<=0xFFFFD \|\|
ali@166	1377	c>=100000 && c<=0x10FFFD)
ali@166	1378	{
ali@166	1379	if (pswit[ECHO_SWITCH])
ali@166	1380	g_print("\n%s\n",aline);
ali@166	1381	if (!pswit[OVERVIEW_SWITCH])
ali@166	1382	g_print(" Line %ld column %ld - Private Use "
ali@166	1383	"character U+%04" G_GINT32_MODIFIER "X\n",
ali@166	1384	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@166	1385	else
ali@166	1386	cnt_bin++;
ali@166	1387	eInvalidChar=TRUE;
ali@166	1388	}
ali@166	1389	}
ali@166	1390	else
ali@166	1391	{
ali@166	1392	t=g_convert_with_iconv(s,g_utf8_next_char(s)-s,
ali@166	1393	charset_validator,NULL,&nb,NULL);
ali@166	1394	if (t)
ali@166	1395	g_free(t);
ali@166	1396	else
ali@166	1397	{
ali@166	1398	if (pswit[ECHO_SWITCH])
ali@166	1399	g_print("\n%s\n",aline);
ali@166	1400	if (!pswit[OVERVIEW_SWITCH])
ali@166	1401	g_print(" Line %ld column %ld - Non-%s "
ali@166	1402	"character %u\n",linecnt,
ali@166	1403	g_utf8_pointer_to_offset(aline,s)+1,charset,c);
ali@166	1404	else
ali@166	1405	cnt_bin++;
ali@166	1406	eInvalidChar=TRUE;
ali@166	1407	}
ali@166	1408	}
ali@44	1409	}
ali@70	1410	if (!eTab && c==CHAR_TAB)
ali@44	1411	{
ali@44	1412	if (pswit[ECHO_SWITCH])
ali@70	1413	g_print("\n%s\n",aline);
ali@44	1414	if (!pswit[OVERVIEW_SWITCH])
ali@70	1415	g_print(" Line %ld column %ld - Tab character?\n",
ali@70	1416	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1417	else
ali@44	1418	cnt_odd++;
ali@70	1419	eTab=TRUE;
ali@44	1420	}
ali@70	1421	if (!eTilde && c==CHAR_TILDE)
ali@44	1422	{
ali@44	1423	/*
ali@44	1424	* Often used by OCR software to indicate an
ali@44	1425	* unrecognizable character.
ali@44	1426	*/
ali@44	1427	if (pswit[ECHO_SWITCH])
ali@70	1428	g_print("\n%s\n",aline);
ali@44	1429	if (!pswit[OVERVIEW_SWITCH])
ali@70	1430	g_print(" Line %ld column %ld - Tilde character?\n",
ali@70	1431	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1432	else
ali@44	1433	cnt_odd++;
ali@70	1434	eTilde=TRUE;
ali@44	1435	}
ali@70	1436	if (!eCarat && c==CHAR_CARAT)
ali@44	1437	{
ali@44	1438	if (pswit[ECHO_SWITCH])
ali@70	1439	g_print("\n%s\n",aline);
ali@44	1440	if (!pswit[OVERVIEW_SWITCH])
ali@70	1441	g_print(" Line %ld column %ld - Carat character?\n",
ali@70	1442	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1443	else
ali@44	1444	cnt_odd++;
ali@70	1445	eCarat=TRUE;
ali@44	1446	}
ali@70	1447	if (!eFSlash && c==CHAR_FORESLASH && warnings->fslash)
ali@44	1448	{
ali@44	1449	if (pswit[ECHO_SWITCH])
ali@70	1450	g_print("\n%s\n",aline);
ali@44	1451	if (!pswit[OVERVIEW_SWITCH])
ali@70	1452	g_print(" Line %ld column %ld - Forward slash?\n",
ali@70	1453	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1454	else
ali@44	1455	cnt_odd++;
ali@70	1456	eFSlash=TRUE;
ali@44	1457	}
ali@44	1458	/*
ali@44	1459	* Report asterisks only in paranoid mode,
ali@44	1460	* since they're often deliberate.
ali@44	1461	*/
ali@44	1462	if (!eAst && pswit[PARANOID_SWITCH] && warnings->ast && !isemptyline &&
ali@70	1463	c==CHAR_ASTERISK)
ali@44	1464	{
ali@44	1465	if (pswit[ECHO_SWITCH])
ali@70	1466	g_print("\n%s\n",aline);
ali@44	1467	if (!pswit[OVERVIEW_SWITCH])
ali@70	1468	g_print(" Line %ld column %ld - Asterisk?\n",
ali@70	1469	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1470	else
ali@44	1471	cnt_odd++;
ali@70	1472	eAst=TRUE;
ali@44	1473	}
ali@44	1474	}
ali@44	1475	}
ali@44	1476
ali@44	1477	/*
ali@45	1478	* check_for_long_line:
ali@45	1479	*
ali@45	1480	* Check for line too long.
ali@45	1481	*/
ali@45	1482	void check_for_long_line(const char *aline)
ali@45	1483	{
ali@70	1484	if (g_utf8_strlen(aline,-1)>LONGEST_PG_LINE)
ali@45	1485	{
ali@45	1486	if (pswit[ECHO_SWITCH])
ali@70	1487	g_print("\n%s\n",aline);
ali@45	1488	if (!pswit[OVERVIEW_SWITCH])
ali@70	1489	g_print(" Line %ld column %ld - Long line %ld\n",
ali@70	1490	linecnt,g_utf8_strlen(aline,-1),g_utf8_strlen(aline,-1));
ali@45	1491	else
ali@45	1492	cnt_long++;
ali@45	1493	}
ali@45	1494	}
ali@45	1495
ali@45	1496	/*
ali@45	1497	* check_for_short_line:
ali@45	1498	*
ali@45	1499	* Check for line too short.
ali@45	1500	*
ali@45	1501	* This one is a bit trickier to implement: we don't want to
ali@45	1502	* flag the last line of a paragraph for being short, so we
ali@45	1503	* have to wait until we know that our current line is a
ali@45	1504	* "normal" line, then report the _previous_ line if it was too
ali@45	1505	* short. We also don't want to report indented lines like
ali@45	1506	* chapter heads or formatted quotations. We therefore keep
ali@45	1507	* last->len as the length of the last line examined, and
ali@45	1508	* last->blen as the length of the last but one, and try to
ali@45	1509	* suppress unnecessary warnings by checking that both were of
ali@45	1510	* "normal" length. We keep the first character of the last
ali@45	1511	* line in last->start, and if it was a space, we assume that
ali@45	1512	* the formatting is deliberate. I can't figure out a way to
ali@45	1513	* distinguish something like a quoted verse left-aligned or
ali@45	1514	* the header or footer of a letter from a paragraph of short
ali@45	1515	* lines - maybe if I examined the whole paragraph, and if the
ali@45	1516	* para has less than, say, 8 lines and if all lines are short,
ali@45	1517	* then just assume it's OK? Need to look at some texts to see
ali@45	1518	* how often a formula like this would get the right result.
ali@45	1519	*/
ali@45	1520	void check_for_short_line(const char aline,const struct line_properties last)
ali@45	1521	{
ali@70	1522	if (g_utf8_strlen(aline,-1)>1 && last->len>1 &&
ali@70	1523	last->len<SHORTEST_PG_LINE && last->blen>1 &&
ali@70	1524	last->blen>SHORTEST_PG_LINE && last->start!=CHAR_SPACE)
ali@45	1525	{
ali@45	1526	if (pswit[ECHO_SWITCH])
ali@70	1527	g_print("\n%s\n",prevline);
ali@45	1528	if (!pswit[OVERVIEW_SWITCH])
ali@70	1529	g_print(" Line %ld column %ld - Short line %ld?\n",
ali@70	1530	linecnt-1,g_utf8_strlen(prevline,-1),g_utf8_strlen(prevline,-1));
ali@45	1531	else
ali@45	1532	cnt_short++;
ali@45	1533	}
ali@45	1534	}
ali@45	1535
ali@45	1536	/*
ali@46	1537	* check_for_starting_punctuation:
ali@46	1538	*
ali@46	1539	* Look for punctuation other than full ellipses at start of line.
ali@46	1540	*/
ali@46	1541	void check_for_starting_punctuation(const char *aline)
ali@46	1542	{
ali@70	1543	if (*aline && g_utf8_strchr(".?!,;:",-1,g_utf8_get_char(aline)) &&
ali@70	1544	!g_str_has_prefix(aline,". . ."))
ali@46	1545	{
ali@46	1546	if (pswit[ECHO_SWITCH])
ali@70	1547	g_print("\n%s\n",aline);
ali@46	1548	if (!pswit[OVERVIEW_SWITCH])
ali@70	1549	g_print(" Line %ld column 1 - Begins with punctuation?\n",
ali@46	1550	linecnt);
ali@46	1551	else
ali@46	1552	cnt_punct++;
ali@46	1553	}
ali@46	1554	}
ali@46	1555
ali@46	1556	/*
ali@147	1557	* str_emdash:
ali@147	1558	*
ali@147	1559	* Find the first em-dash, return a pointer to it and set <next> to the
ali@147	1560	* character following the dash.
ali@147	1561	*/
ali@147	1562	char str_emdash(const char s,const char **next)
ali@147	1563	{
ali@147	1564	const char s1,s2;
ali@147	1565	s1=strstr(s,"--");
ali@147	1566	s2=strstr(s,"—");
ali@147	1567	if (!s1)
ali@147	1568	{
ali@147	1569	if (s2)
ali@147	1570	*next=g_utf8_next_char(s2);
ali@147	1571	return (char *)s2;
ali@147	1572	}
ali@147	1573	else if (!s2)
ali@147	1574	{
ali@147	1575	*next=g_utf8_next_char(g_utf8_next_char(s1));
ali@147	1576	return (char *)s1;
ali@147	1577	}
ali@147	1578	else if (s1<s2)
ali@147	1579	{
ali@147	1580	*next=g_utf8_next_char(g_utf8_next_char(s1));
ali@147	1581	return (char *)s1;
ali@147	1582	}
ali@147	1583	else
ali@147	1584	{
ali@147	1585	*next=g_utf8_next_char(s2);
ali@147	1586	return (char *)s2;
ali@147	1587	}
ali@147	1588	}
ali@147	1589
ali@147	1590	/*
ali@47	1591	* check_for_spaced_emdash:
ali@47	1592	*
ali@47	1593	* Check for spaced em-dashes.
ali@47	1594	*
ali@147	1595	* We must check _all_ occurrences of em-dashes on the line
ali@147	1596	* hence the loop - even if the first dash is OK
ali@47	1597	* there may be another that's wrong later on.
ali@47	1598	*/
ali@47	1599	void check_for_spaced_emdash(const char *aline)
ali@47	1600	{
ali@70	1601	const char s,t,*next;
ali@147	1602	for (s=aline;t=str_emdash(s,&next);s=next)
ali@47	1603	{
ali@70	1604	if (t>aline && g_utf8_get_char(g_utf8_prev_char(t))==CHAR_SPACE \|\|
ali@70	1605	g_utf8_get_char(next)==CHAR_SPACE)
ali@47	1606	{
ali@47	1607	if (pswit[ECHO_SWITCH])
ali@70	1608	g_print("\n%s\n",aline);
ali@47	1609	if (!pswit[OVERVIEW_SWITCH])
ali@70	1610	g_print(" Line %ld column %ld - Spaced em-dash?\n",
ali@70	1611	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@47	1612	else
ali@47	1613	cnt_dash++;
ali@47	1614	}
ali@47	1615	}
ali@47	1616	}
ali@47	1617
ali@47	1618	/*
ali@47	1619	* check_for_spaced_dash:
ali@47	1620	*
ali@47	1621	* Check for spaced dashes.
ali@47	1622	*/
ali@47	1623	void check_for_spaced_dash(const char *aline)
ali@47	1624	{
ali@47	1625	const char *s;
ali@47	1626	if ((s=strstr(aline," -")))
ali@47	1627	{
ali@70	1628	if (g_utf8_get_char(g_utf8_next_char(g_utf8_next_char(s)))!='-')
ali@47	1629	{
ali@47	1630	if (pswit[ECHO_SWITCH])
ali@70	1631	g_print("\n%s\n",aline);
ali@47	1632	if (!pswit[OVERVIEW_SWITCH])
ali@70	1633	g_print(" Line %ld column %ld - Spaced dash?\n",
ali@70	1634	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@47	1635	else
ali@47	1636	cnt_dash++;
ali@47	1637	}
ali@47	1638	}
ali@47	1639	else if ((s=strstr(aline,"- ")))
ali@47	1640	{
ali@70	1641	if (s==aline \|\| g_utf8_get_char(g_utf8_prev_char(s))!='-')
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 dash?\n",
ali@70	1647	linecnt,g_utf8_pointer_to_offset(aline,s)+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@48	1655	* check_for_unmarked_paragraphs:
ali@48	1656	*
ali@48	1657	* Check for unmarked paragraphs indicated by separate speakers.
ali@48	1658	*
ali@48	1659	* May well be false positive:
ali@48	1660	* "Bravo!" "Wonderful!" called the crowd.
ali@48	1661	* but useful all the same.
ali@48	1662	*/
ali@48	1663	void check_for_unmarked_paragraphs(const char *aline)
ali@48	1664	{
ali@48	1665	const char *s;
ali@48	1666	s=strstr(aline,"\" \"");
ali@48	1667	if (!s)
ali@48	1668	s=strstr(aline,"\" \"");
ali@48	1669	if (s)
ali@48	1670	{
ali@48	1671	if (pswit[ECHO_SWITCH])
ali@70	1672	g_print("\n%s\n",aline);
ali@48	1673	if (!pswit[OVERVIEW_SWITCH])
ali@70	1674	g_print(" Line %ld column %ld - "
ali@70	1675	"Query missing paragraph break?\n",
ali@70	1676	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@48	1677	else
ali@48	1678	cnt_punct++;
ali@48	1679	}
ali@48	1680	}
ali@48	1681
ali@48	1682	/*
ali@49	1683	* check_for_jeebies:
ali@49	1684	*
ali@49	1685	* Check for "to he" and other easy h/b errors.
ali@49	1686	*
ali@49	1687	* This is a very inadequate effort on the h/b problem,
ali@49	1688	* but the phrase "to he" is always an error, whereas "to
ali@49	1689	* be" is quite common.
ali@49	1690	* Similarly, '"Quiet!", be said.' is a non-be error
ali@49	1691	* "to he" is _not_ always an error!:
ali@49	1692	* "Where they went to he couldn't say."
ali@49	1693	* Another false positive:
ali@49	1694	* What would "Cinderella" be without the . . .
ali@49	1695	* and another: "If he wants to he can see for himself."
ali@49	1696	*/
ali@49	1697	void check_for_jeebies(const char *aline)
ali@49	1698	{
ali@49	1699	const char *s;
ali@49	1700	s=strstr(aline," be could ");
ali@49	1701	if (!s)
ali@49	1702	s=strstr(aline," be would ");
ali@49	1703	if (!s)
ali@49	1704	s=strstr(aline," was be ");
ali@49	1705	if (!s)
ali@49	1706	s=strstr(aline," be is ");
ali@49	1707	if (!s)
ali@49	1708	s=strstr(aline," is be ");
ali@49	1709	if (!s)
ali@49	1710	s=strstr(aline,"\", be ");
ali@49	1711	if (!s)
ali@49	1712	s=strstr(aline,"\" be ");
ali@49	1713	if (!s)
ali@49	1714	s=strstr(aline,"\" be ");
ali@49	1715	if (!s)
ali@49	1716	s=strstr(aline," to he ");
ali@49	1717	if (s)
ali@49	1718	{
ali@49	1719	if (pswit[ECHO_SWITCH])
ali@70	1720	g_print("\n%s\n",aline);
ali@49	1721	if (!pswit[OVERVIEW_SWITCH])
ali@70	1722	g_print(" Line %ld column %ld - Query he/be error?\n",
ali@70	1723	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@49	1724	else
ali@49	1725	cnt_word++;
ali@49	1726	}
ali@49	1727	s=strstr(aline," the had ");
ali@49	1728	if (!s)
ali@49	1729	s=strstr(aline," a had ");
ali@49	1730	if (!s)
ali@49	1731	s=strstr(aline," they bad ");
ali@49	1732	if (!s)
ali@49	1733	s=strstr(aline," she bad ");
ali@49	1734	if (!s)
ali@49	1735	s=strstr(aline," he bad ");
ali@49	1736	if (!s)
ali@49	1737	s=strstr(aline," you bad ");
ali@49	1738	if (!s)
ali@49	1739	s=strstr(aline," i bad ");
ali@49	1740	if (s)
ali@49	1741	{
ali@49	1742	if (pswit[ECHO_SWITCH])
ali@70	1743	g_print("\n%s\n",aline);
ali@49	1744	if (!pswit[OVERVIEW_SWITCH])
ali@70	1745	g_print(" Line %ld column %ld - Query had/bad error?\n",
ali@70	1746	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@49	1747	else
ali@49	1748	cnt_word++;
ali@49	1749	}
ali@49	1750	s=strstr(aline,"; hut ");
ali@49	1751	if (!s)
ali@49	1752	s=strstr(aline,", hut ");
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 hut/but 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	}
ali@49	1764
ali@49	1765	/*
ali@50	1766	* check_for_mta_from:
ali@50	1767	*
ali@50	1768	* Special case - angled bracket in front of "From" placed there by an
ali@50	1769	* MTA when sending an e-mail.
ali@50	1770	*/
ali@50	1771	void check_for_mta_from(const char *aline)
ali@50	1772	{
ali@50	1773	const char *s;
ali@50	1774	s=strstr(aline,">From");
ali@50	1775	if (s)
ali@50	1776	{
ali@50	1777	if (pswit[ECHO_SWITCH])
ali@70	1778	g_print("\n%s\n",aline);
ali@50	1779	if (!pswit[OVERVIEW_SWITCH])
ali@70	1780	g_print(" Line %ld column %ld - "
ali@70	1781	"Query angled bracket with From\n",
ali@70	1782	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@50	1783	else
ali@50	1784	cnt_punct++;
ali@50	1785	}
ali@50	1786	}
ali@50	1787
ali@50	1788	/*
ali@51	1789	* check_for_orphan_character:
ali@51	1790	*
ali@51	1791	* Check for a single character line -
ali@51	1792	* often an overflow from bad wrapping.
ali@51	1793	*/
ali@51	1794	void check_for_orphan_character(const char *aline)
ali@51	1795	{
ali@70	1796	gunichar c;
ali@70	1797	c=g_utf8_get_char(aline);
ali@70	1798	if (c && !*g_utf8_next_char(aline))
ali@51	1799	{
ali@70	1800	if (c=='I' \|\| c=='V' \|\| c=='X' \|\| c=='L' \|\| g_unichar_isdigit(c))
ali@51	1801	; /* Nothing - ignore numerals alone on a line. */
ali@51	1802	else
ali@51	1803	{
ali@51	1804	if (pswit[ECHO_SWITCH])
ali@70	1805	g_print("\n%s\n",aline);
ali@51	1806	if (!pswit[OVERVIEW_SWITCH])
ali@70	1807	g_print(" Line %ld column 1 - Query single character line\n",
ali@51	1808	linecnt);
ali@51	1809	else
ali@51	1810	cnt_punct++;
ali@51	1811	}
ali@51	1812	}
ali@51	1813	}
ali@51	1814
ali@51	1815	/*
ali@52	1816	* check_for_pling_scanno:
ali@52	1817	*
ali@52	1818	* Check for I" - often should be !
ali@52	1819	*/
ali@52	1820	void check_for_pling_scanno(const char *aline)
ali@52	1821	{
ali@52	1822	const char *s;
ali@52	1823	s=strstr(aline," I\"");
ali@52	1824	if (s)
ali@52	1825	{
ali@52	1826	if (pswit[ECHO_SWITCH])
ali@70	1827	g_print("\n%s\n",aline);
ali@52	1828	if (!pswit[OVERVIEW_SWITCH])
ali@70	1829	g_print(" Line %ld column %ld - Query I=exclamation mark?\n",
ali@70	1830	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@52	1831	else
ali@52	1832	cnt_punct++;
ali@52	1833	}
ali@52	1834	}
ali@52	1835
ali@52	1836	/*
ali@53	1837	* check_for_extra_period:
ali@53	1838	*
ali@53	1839	* Check for period without a capital letter. Cut-down from gutspell.
ali@53	1840	* Only works when it happens on a single line.
ali@53	1841	*/
ali@53	1842	void check_for_extra_period(const char aline,const struct warnings warnings)
ali@53	1843	{
ali@99	1844	const char s,t,s1,sprev;
ali@69	1845	int i;
ali@70	1846	gsize len;
ali@69	1847	gboolean istypo;
ali@69	1848	gchar *testword;
ali@99	1849	gunichar c,nc,pc,*decomposition;
ali@53	1850	if (pswit[PARANOID_SWITCH])
ali@53	1851	{
ali@70	1852	for (t=aline;t=strstr(t,". ");)
ali@53	1853	{
ali@69	1854	if (t==aline)
ali@53	1855	{
ali@70	1856	t=g_utf8_next_char(t);
ali@53	1857	/* start of line punctuation is handled elsewhere */
ali@53	1858	continue;
ali@53	1859	}
ali@70	1860	if (!g_unichar_isalpha(g_utf8_get_char(g_utf8_prev_char(t))))
ali@53	1861	{
ali@70	1862	t=g_utf8_next_char(t);
ali@53	1863	continue;
ali@53	1864	}
ali@53	1865	if (warnings->isDutch)
ali@53	1866	{
ali@53	1867	/* For Frank & Jeroen -- 's Middags case */
ali@70	1868	gunichar c2,c3,c4,c5;
ali@70	1869	c2=g_utf8_get_char(g_utf8_offset_to_pointer(t,2));
ali@70	1870	c3=g_utf8_get_char(g_utf8_offset_to_pointer(t,3));
ali@70	1871	c4=g_utf8_get_char(g_utf8_offset_to_pointer(t,4));
ali@70	1872	c5=g_utf8_get_char(g_utf8_offset_to_pointer(t,5));
ali@99	1873	if (CHAR_IS_APOSTROPHE(c2) &&
ali@99	1874	g_unichar_islower(c3) && c4==CHAR_SPACE &&
ali@99	1875	g_unichar_isupper(c5))
ali@53	1876	{
ali@70	1877	t=g_utf8_next_char(t);
ali@53	1878	continue;
ali@53	1879	}
ali@53	1880	}
ali@70	1881	s1=g_utf8_next_char(g_utf8_next_char(t));
ali@70	1882	while (*s1 && !g_unichar_isalpha(g_utf8_get_char(s1)) &&
ali@70	1883	!isdigit(g_utf8_get_char(s1)))
ali@70	1884	s1=g_utf8_next_char(s1);
ali@70	1885	if (g_unichar_islower(g_utf8_get_char(s1)))
ali@53	1886	{
ali@53	1887	/* we have something to investigate */
ali@69	1888	istypo=TRUE;
ali@53	1889	/* so let's go back and find out */
ali@99	1890	nc=g_utf8_get_char(t);
ali@99	1891	s1=g_utf8_prev_char(t);
ali@99	1892	c=g_utf8_get_char(s1);
ali@99	1893	sprev=g_utf8_prev_char(s1);
ali@99	1894	pc=g_utf8_get_char(sprev);
ali@99	1895	while (s1>=aline &&
ali@99	1896	(g_unichar_isalpha(c) \|\| g_unichar_isdigit(c) \|\|
ali@99	1897	g_unichar_isalpha(pc) && CHAR_IS_APOSTROPHE(c) &&
ali@99	1898	g_unichar_isalpha(nc)))
ali@99	1899	{
ali@99	1900	nc=c;
ali@99	1901	s1=sprev;
ali@99	1902	c=pc;
ali@99	1903	sprev=g_utf8_prev_char(s1);
ali@99	1904	pc=g_utf8_get_char(sprev);
ali@99	1905	}
ali@70	1906	s1=g_utf8_next_char(s1);
ali@69	1907	s=strchr(s1,'.');
ali@69	1908	if (s)
ali@69	1909	testword=g_strndup(s1,s-s1);
ali@69	1910	else
ali@69	1911	testword=g_strdup(s1);
ali@53	1912	for (i=0;*abbrev[i];i++)
ali@53	1913	if (!strcmp(testword,abbrev[i]))
ali@69	1914	istypo=FALSE;
ali@70	1915	if (g_unichar_isdigit(g_utf8_get_char(testword)))
ali@69	1916	istypo=FALSE;
ali@70	1917	if (!*g_utf8_next_char(testword))
ali@69	1918	istypo=FALSE;
ali@53	1919	if (isroman(testword))
ali@69	1920	istypo=FALSE;
ali@53	1921	if (istypo)
ali@53	1922	{
ali@69	1923	istypo=FALSE;
ali@70	1924	for (s=testword;*s;s=g_utf8_next_char(s))
ali@70	1925	{
ali@70	1926	decomposition=g_unicode_canonical_decomposition(
ali@70	1927	g_utf8_get_char(s),&len);
ali@70	1928	if (g_utf8_strchr("aeiou",-1,decomposition[0]))
ali@69	1929	istypo=TRUE;
ali@70	1930	g_free(decomposition);
ali@70	1931	}
ali@53	1932	}
ali@69	1933	if (istypo &&
ali@69	1934	(pswit[VERBOSE_SWITCH] \|\| !g_tree_lookup(qperiod,testword)))
ali@53	1935	{
ali@69	1936	g_tree_insert(qperiod,g_strdup(testword),
ali@69	1937	GINT_TO_POINTER(1));
ali@69	1938	if (pswit[ECHO_SWITCH])
ali@70	1939	g_print("\n%s\n",aline);
ali@69	1940	if (!pswit[OVERVIEW_SWITCH])
ali@70	1941	g_print(" Line %ld column %ld - Extra period?\n",
ali@70	1942	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@69	1943	else
ali@69	1944	cnt_punct++;
ali@53	1945	}
ali@69	1946	g_free(testword);
ali@53	1947	}
ali@70	1948	t=g_utf8_next_char(t);
ali@53	1949	}
ali@53	1950	}
ali@53	1951	}
ali@53	1952
ali@53	1953	/*
ali@54	1954	* check_for_following_punctuation:
ali@54	1955	*
ali@54	1956	* Check for words usually not followed by punctuation.
ali@54	1957	*/
ali@54	1958	void check_for_following_punctuation(const char *aline)
ali@54	1959	{
ali@54	1960	int i;
ali@54	1961	const char s,wordstart;
ali@70	1962	gunichar c;
ali@69	1963	gchar inword,t;
ali@54	1964	if (pswit[TYPO_SWITCH])
ali@54	1965	{
ali@54	1966	for (s=aline;*s;)
ali@54	1967	{
ali@54	1968	wordstart=s;
ali@69	1969	t=getaword(&s);
ali@69	1970	if (!*t)
ali@69	1971	{
ali@69	1972	g_free(t);
ali@54	1973	continue;
ali@69	1974	}
ali@70	1975	inword=g_utf8_strdown(t,-1);
ali@69	1976	g_free(t);
ali@54	1977	for (i=0;*nocomma[i];i++)
ali@54	1978	if (!strcmp(inword,nocomma[i]))
ali@54	1979	{
ali@70	1980	c=g_utf8_get_char(s);
ali@70	1981	if (c==',' \|\| c==';' \|\| c==':')
ali@54	1982	{
ali@54	1983	if (pswit[ECHO_SWITCH])
ali@70	1984	g_print("\n%s\n",aline);
ali@54	1985	if (!pswit[OVERVIEW_SWITCH])
ali@70	1986	g_print(" Line %ld column %ld - "
ali@54	1987	"Query punctuation after %s?\n",
ali@70	1988	linecnt,g_utf8_pointer_to_offset(aline,s)+1,
ali@70	1989	inword);
ali@54	1990	else
ali@54	1991	cnt_punct++;
ali@54	1992	}
ali@54	1993	}
ali@54	1994	for (i=0;*noperiod[i];i++)
ali@54	1995	if (!strcmp(inword,noperiod[i]))
ali@54	1996	{
ali@70	1997	c=g_utf8_get_char(s);
ali@70	1998	if (c=='.' \|\| c=='!')
ali@54	1999	{
ali@54	2000	if (pswit[ECHO_SWITCH])
ali@70	2001	g_print("\n%s\n",aline);
ali@54	2002	if (!pswit[OVERVIEW_SWITCH])
ali@70	2003	g_print(" Line %ld column %ld - "
ali@54	2004	"Query punctuation after %s?\n",
ali@70	2005	linecnt,g_utf8_pointer_to_offset(aline,s)+1,
ali@70	2006	inword);
ali@54	2007	else
ali@54	2008	cnt_punct++;
ali@54	2009	}
ali@54	2010	}
ali@69	2011	g_free(inword);
ali@54	2012	}
ali@54	2013	}
ali@54	2014	}
ali@54	2015
ali@54	2016	/*
ali@55	2017	* check_for_typos:
ali@55	2018	*
ali@55	2019	* Check for commonly mistyped words,
ali@55	2020	* and digits like 0 for O in a word.
ali@55	2021	*/
ali@55	2022	void check_for_typos(const char aline,struct warnings warnings)
ali@55	2023	{
ali@70	2024	const char s,t,nt,wordstart;
ali@70	2025	gchar *inword;
ali@70	2026	gunichar *decomposition;
ali@70	2027	gchar *testword;
ali@70	2028	int i,vowel,consonant,*dupcnt;
ali@70	2029	gboolean isdup,istypo,alower;
ali@99	2030	gunichar c,pc;
ali@70	2031	long offset,len;
ali@70	2032	gsize decomposition_len;
ali@55	2033	for (s=aline;*s;)
ali@55	2034	{
ali@55	2035	wordstart=s;
ali@69	2036	inword=getaword(&s);
ali@55	2037	if (!*inword)
ali@69	2038	{
ali@69	2039	g_free(inword);
ali@55	2040	continue; /* don't bother with empty lines */
ali@69	2041	}
ali@55	2042	if (mixdigit(inword))
ali@55	2043	{
ali@55	2044	if (pswit[ECHO_SWITCH])
ali@70	2045	g_print("\n%s\n",aline);
ali@55	2046	if (!pswit[OVERVIEW_SWITCH])
ali@70	2047	g_print(" Line %ld column %ld - Query digit in %s\n",
ali@70	2048	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+1,inword);
ali@55	2049	else
ali@55	2050	cnt_word++;
ali@55	2051	}
ali@55	2052	/*
ali@55	2053	* Put the word through a series of tests for likely typos and OCR
ali@55	2054	* errors.
ali@55	2055	*/
ali@69	2056	if (pswit[TYPO_SWITCH] \|\| pswit[USERTYPO_SWITCH])
ali@55	2057	{
ali@69	2058	istypo=FALSE;
ali@70	2059	alower=FALSE;
ali@70	2060	for (t=inword;*t;t=g_utf8_next_char(t))
ali@55	2061	{
ali@70	2062	c=g_utf8_get_char(t);
ali@70	2063	nt=g_utf8_next_char(t);
ali@55	2064	/* lowercase for testing */
ali@70	2065	if (g_unichar_islower(c))
ali@70	2066	alower=TRUE;
ali@70	2067	if (alower && (g_unichar_isupper(c) \|\| g_unichar_istitle(c)))
ali@55	2068	{
ali@55	2069	/*
ali@55	2070	* We have an uppercase mid-word. However, there are
ali@55	2071	* common cases:
ali@55	2072	* Mac and Mc like McGill
ali@55	2073	* French contractions like l'Abbe
ali@55	2074	*/
ali@70	2075	offset=g_utf8_pointer_to_offset(inword,t);
ali@99	2076	if (offset>0)
ali@99	2077	pc=g_utf8_get_char(g_utf8_prev_char(t));
ali@99	2078	else
ali@99	2079	pc='\0';
ali@70	2080	if (offset==2 && c=='m' && g_utf8_get_char(nt)=='c' \|\|
ali@70	2081	offset==3 && c=='m' && g_utf8_get_char(nt)=='a' &&
ali@70	2082	g_utf8_get_char(g_utf8_next_char(nt))=='c' \|\|
ali@99	2083	CHAR_IS_APOSTROPHE(pc))
ali@55	2084	; /* do nothing! */
ali@55	2085	else
ali@69	2086	istypo=TRUE;
ali@55	2087	}
ali@55	2088	}
ali@70	2089	testword=g_utf8_casefold(inword,-1);
ali@69	2090	}
ali@69	2091	if (pswit[TYPO_SWITCH])
ali@69	2092	{
ali@55	2093	/*
ali@55	2094	* Check for certain unlikely two-letter combinations at word
ali@55	2095	* start and end.
ali@55	2096	*/
ali@70	2097	len=g_utf8_strlen(testword,-1);
ali@70	2098	if (len>1)
ali@55	2099	{
ali@55	2100	for (i=0;*nostart[i];i++)
ali@70	2101	if (g_str_has_prefix(testword,nostart[i]))
ali@69	2102	istypo=TRUE;
ali@55	2103	for (i=0;*noend[i];i++)
ali@70	2104	if (g_str_has_suffix(testword,noend[i]))
ali@69	2105	istypo=TRUE;
ali@55	2106	}
ali@55	2107	/* ght is common, gbt never. Like that. */
ali@55	2108	if (strstr(testword,"cb"))
ali@69	2109	istypo=TRUE;
ali@55	2110	if (strstr(testword,"gbt"))
ali@69	2111	istypo=TRUE;
ali@55	2112	if (strstr(testword,"pbt"))
ali@69	2113	istypo=TRUE;
ali@55	2114	if (strstr(testword,"tbs"))
ali@69	2115	istypo=TRUE;
ali@55	2116	if (strstr(testword,"mrn"))
ali@69	2117	istypo=TRUE;
ali@55	2118	if (strstr(testword,"ahle"))
ali@69	2119	istypo=TRUE;
ali@55	2120	if (strstr(testword,"ihle"))
ali@69	2121	istypo=TRUE;
ali@55	2122	/*
ali@55	2123	* "TBE" does happen - like HEARTBEAT - but uncommon.
ali@55	2124	* Also "TBI" - frostbite, outbid - but uncommon.
ali@55	2125	* Similarly "ii" like Hawaii, or Pompeii, and in Roman
ali@55	2126	* numerals, but "ii" is a common scanno.
ali@55	2127	*/
ali@55	2128	if (strstr(testword,"tbi"))
ali@69	2129	istypo=TRUE;
ali@55	2130	if (strstr(testword,"tbe"))
ali@69	2131	istypo=TRUE;
ali@55	2132	if (strstr(testword,"ii"))
ali@69	2133	istypo=TRUE;
ali@55	2134	/*
ali@55	2135	* Check for no vowels or no consonants.
ali@55	2136	* If none, flag a typo.
ali@55	2137	*/
ali@70	2138	if (!istypo && len>1)
ali@55	2139	{
ali@55	2140	vowel=consonant=0;
ali@70	2141	for (t=testword;*t;t=g_utf8_next_char(t))
ali@55	2142	{
ali@70	2143	c=g_utf8_get_char(t);
ali@70	2144	decomposition=
ali@70	2145	g_unicode_canonical_decomposition(c,&decomposition_len);
ali@70	2146	if (c=='y' \|\| g_unichar_isdigit(c))
ali@55	2147	{
ali@55	2148	/* Yah, this is loose. */
ali@55	2149	vowel++;
ali@55	2150	consonant++;
ali@55	2151	}
ali@70	2152	else if (g_utf8_strchr("aeiou",-1,decomposition[0]))
ali@55	2153	vowel++;
ali@55	2154	else
ali@55	2155	consonant++;
ali@70	2156	g_free(decomposition);
ali@55	2157	}
ali@55	2158	if (!vowel \|\| !consonant)
ali@69	2159	istypo=TRUE;
ali@55	2160	}
ali@55	2161	/*
ali@55	2162	* Now exclude the word from being reported if it's in
ali@55	2163	* the okword list.
ali@55	2164	*/
ali@55	2165	for (i=0;*okword[i];i++)
ali@55	2166	if (!strcmp(testword,okword[i]))
ali@69	2167	istypo=FALSE;
ali@55	2168	/*
ali@55	2169	* What looks like a typo may be a Roman numeral.
ali@55	2170	* Exclude these.
ali@55	2171	*/
ali@55	2172	if (istypo && isroman(testword))
ali@69	2173	istypo=FALSE;
ali@55	2174	/* Check the manual list of typos. */
ali@55	2175	if (!istypo)
ali@55	2176	for (i=0;*typo[i];i++)
ali@55	2177	if (!strcmp(testword,typo[i]))
ali@69	2178	istypo=TRUE;
ali@55	2179	/*
ali@55	2180	* Check lowercase s, l, i and m - special cases.
ali@55	2181	* "j" - often a semi-colon gone wrong.
ali@55	2182	* "d" for a missing apostrophe - he d
ali@55	2183	* "n" for "in"
ali@55	2184	*/
ali@70	2185	if (!istypo && len==1 &&
ali@70	2186	g_utf8_strchr("slmijdn",-1,g_utf8_get_char(inword)))
ali@69	2187	istypo=TRUE;
ali@55	2188	if (istypo)
ali@55	2189	{
ali@69	2190	dupcnt=g_tree_lookup(qword,testword);
ali@69	2191	if (dupcnt)
ali@69	2192	{
ali@69	2193	(*dupcnt)++;
ali@69	2194	isdup=!pswit[VERBOSE_SWITCH];
ali@69	2195	}
ali@69	2196	else
ali@69	2197	{
ali@69	2198	dupcnt=g_new0(int,1);
ali@69	2199	g_tree_insert(qword,g_strdup(testword),dupcnt);
ali@69	2200	isdup=FALSE;
ali@69	2201	}
ali@55	2202	if (!isdup)
ali@55	2203	{
ali@55	2204	if (pswit[ECHO_SWITCH])
ali@70	2205	g_print("\n%s\n",aline);
ali@55	2206	if (!pswit[OVERVIEW_SWITCH])
ali@55	2207	{
ali@70	2208	g_print(" Line %ld column %ld - Query word %s",
ali@70	2209	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+1,
ali@70	2210	inword);
ali@69	2211	if (!pswit[VERBOSE_SWITCH])
ali@70	2212	g_print(" - not reporting duplicates");
ali@70	2213	g_print("\n");
ali@55	2214	}
ali@55	2215	else
ali@55	2216	cnt_word++;
ali@55	2217	}
ali@55	2218	}
ali@55	2219	}
ali@55	2220	/* check the user's list of typos */
ali@69	2221	if (!istypo && usertypo && g_tree_lookup(usertypo,testword))
ali@69	2222	{
ali@69	2223	if (pswit[ECHO_SWITCH])
ali@70	2224	g_print("\n%s\n",aline);
ali@69	2225	if (!pswit[OVERVIEW_SWITCH])
ali@70	2226	g_print(" Line %ld column %ld - Query possible scanno %s\n",
ali@70	2227	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+2,inword);
ali@69	2228	}
ali@69	2229	if (pswit[TYPO_SWITCH] \|\| pswit[USERTYPO_SWITCH])
ali@69	2230	g_free(testword);
ali@55	2231	if (pswit[PARANOID_SWITCH] && warnings->digit)
ali@55	2232	{
ali@55	2233	/* In paranoid mode, query all 0 and 1 standing alone. */
ali@55	2234	if (!strcmp(inword,"0") \|\| !strcmp(inword,"1"))
ali@55	2235	{
ali@55	2236	if (pswit[ECHO_SWITCH])
ali@70	2237	g_print("\n%s\n",aline);
ali@55	2238	if (!pswit[OVERVIEW_SWITCH])
ali@70	2239	g_print(" Line %ld column %ld - Query standalone %s\n",
ali@70	2240	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+2,
ali@70	2241	inword);
ali@55	2242	else
ali@55	2243	cnt_word++;
ali@55	2244	}
ali@55	2245	}
ali@69	2246	g_free(inword);
ali@55	2247	}
ali@55	2248	}
ali@55	2249
ali@56	2250	/*
ali@56	2251	* check_for_misspaced_punctuation:
ali@56	2252	*
ali@56	2253	* Look for added or missing spaces around punctuation and quotes.
ali@56	2254	* If there is a punctuation character like ! with no space on
ali@56	2255	* either side, suspect a missing!space. If there are spaces on
ali@56	2256	* both sides , assume a typo. If we see a double quote with no
ali@56	2257	* space or punctuation on either side of it, assume unspaced
ali@56	2258	* quotes "like"this.
ali@56	2259	*/
ali@56	2260	void check_for_misspaced_punctuation(const char *aline,
ali@69	2261	struct parities *parities,gboolean isemptyline)
ali@56	2262	{
ali@69	2263	gboolean isacro,isellipsis;
ali@56	2264	const char *s;
ali@70	2265	gunichar c,nc,pc,n2c;
ali@142	2266	int parity;
ali@70	2267	c=g_utf8_get_char(aline);
ali@70	2268	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2269	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2270	{
ali@70	2271	pc=c;
ali@70	2272	c=nc;
ali@70	2273	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2274	/* For each character in the line after the first. */
ali@70	2275	if (g_utf8_strchr(".?!,;:_",-1,c)) /* if it's punctuation */
ali@56	2276	{
ali@56	2277	/* we need to suppress warnings for acronyms like M.D. */
ali@69	2278	isacro=FALSE;
ali@56	2279	/* we need to suppress warnings for ellipsis . . . */
ali@69	2280	isellipsis=FALSE;
ali@70	2281	/*
ali@70	2282	* If there are letters on both sides of it or
ali@70	2283	* if it's strict punctuation followed by an alpha.
ali@70	2284	*/
ali@70	2285	if (g_unichar_isalpha(nc) && (g_unichar_isalpha(pc) \|\|
ali@70	2286	g_utf8_strchr("?!,;:",-1,c)))
ali@56	2287	{
ali@70	2288	if (c=='.')
ali@56	2289	{
ali@70	2290	if (g_utf8_pointer_to_offset(aline,s)>2 &&
ali@70	2291	g_utf8_get_char(g_utf8_offset_to_pointer(s,-2))=='.')
ali@69	2292	isacro=TRUE;
ali@70	2293	n2c=g_utf8_get_char(g_utf8_next_char(g_utf8_next_char(s)));
ali@70	2294	if (nc && n2c=='.')
ali@69	2295	isacro=TRUE;
ali@56	2296	}
ali@56	2297	if (!isacro)
ali@56	2298	{
ali@56	2299	if (pswit[ECHO_SWITCH])
ali@70	2300	g_print("\n%s\n",aline);
ali@56	2301	if (!pswit[OVERVIEW_SWITCH])
ali@70	2302	g_print(" Line %ld column %ld - Missing space?\n",
ali@70	2303	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2304	else
ali@56	2305	cnt_punct++;
ali@56	2306	}
ali@56	2307	}
ali@70	2308	if (pc==CHAR_SPACE && (nc==CHAR_SPACE \|\| !nc))
ali@56	2309	{
ali@56	2310	/*
ali@56	2311	* If there are spaces on both sides,
ali@56	2312	* or space before and end of line.
ali@56	2313	*/
ali@70	2314	if (c=='.')
ali@56	2315	{
ali@70	2316	if (g_utf8_pointer_to_offset(aline,s)>2 &&
ali@70	2317	g_utf8_get_char(g_utf8_offset_to_pointer(s,-2))=='.')
ali@69	2318	isellipsis=TRUE;
ali@70	2319	n2c=g_utf8_get_char(g_utf8_next_char(g_utf8_next_char(s)));
ali@70	2320	if (nc && n2c=='.')
ali@69	2321	isellipsis=TRUE;
ali@56	2322	}
ali@56	2323	if (!isemptyline && !isellipsis)
ali@56	2324	{
ali@56	2325	if (pswit[ECHO_SWITCH])
ali@70	2326	g_print("\n%s\n",aline);
ali@56	2327	if (!pswit[OVERVIEW_SWITCH])
ali@70	2328	g_print(" Line %ld column %ld - "
ali@70	2329	"Spaced punctuation?\n",linecnt,
ali@70	2330	g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2331	else
ali@56	2332	cnt_punct++;
ali@56	2333	}
ali@56	2334	}
ali@56	2335	}
ali@56	2336	}
ali@56	2337	/* Split out the characters that CANNOT be preceded by space. */
ali@70	2338	c=g_utf8_get_char(aline);
ali@70	2339	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2340	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2341	{
ali@70	2342	pc=c;
ali@70	2343	c=nc;
ali@70	2344	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2345	/* for each character in the line after the first */
ali@70	2346	if (g_utf8_strchr("?!,;:",-1,c))
ali@56	2347	{
ali@56	2348	/* if it's punctuation that _cannot_ have a space before it */
ali@70	2349	if (pc==CHAR_SPACE && !isemptyline && nc!=CHAR_SPACE)
ali@56	2350	{
ali@56	2351	/*
ali@70	2352	* If nc DOES == space,
ali@56	2353	* it was already reported just above.
ali@56	2354	*/
ali@56	2355	if (pswit[ECHO_SWITCH])
ali@70	2356	g_print("\n%s\n",aline);
ali@56	2357	if (!pswit[OVERVIEW_SWITCH])
ali@70	2358	g_print(" Line %ld column %ld - Spaced punctuation?\n",
ali@70	2359	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2360	else
ali@56	2361	cnt_punct++;
ali@56	2362	}
ali@56	2363	}
ali@56	2364	}
ali@56	2365	/*
ali@56	2366	* Special case " .X" where X is any alpha.
ali@56	2367	* This plugs a hole in the acronym code above.
ali@56	2368	* Inelegant, but maintainable.
ali@56	2369	*/
ali@70	2370	c=g_utf8_get_char(aline);
ali@70	2371	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2372	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2373	{
ali@70	2374	pc=c;
ali@70	2375	c=nc;
ali@70	2376	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2377	/* for each character in the line after the first */
ali@70	2378	if (c=='.')
ali@56	2379	{
ali@56	2380	/* if it's a period */
ali@70	2381	if (pc==CHAR_SPACE && g_unichar_isalpha(nc))
ali@56	2382	{
ali@56	2383	/*
ali@56	2384	* If the period follows a space and
ali@56	2385	* is followed by a letter.
ali@56	2386	*/
ali@56	2387	if (pswit[ECHO_SWITCH])
ali@70	2388	g_print("\n%s\n",aline);
ali@56	2389	if (!pswit[OVERVIEW_SWITCH])
ali@70	2390	g_print(" Line %ld column %ld - Spaced punctuation?\n",
ali@70	2391	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2392	else
ali@56	2393	cnt_punct++;
ali@56	2394	}
ali@56	2395	}
ali@56	2396	}
ali@70	2397	c=g_utf8_get_char(aline);
ali@70	2398	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2399	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2400	{
ali@70	2401	pc=c;
ali@70	2402	c=nc;
ali@70	2403	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2404	/* for each character in the line after the first */
ali@142	2405	if (CHAR_IS_DQUOTE(c))
ali@56	2406	{
ali@70	2407	if (!g_utf8_strchr(" _-.'`,;:!/([{?}])",-1,pc) &&
ali@70	2408	!g_utf8_strchr(" _-.'`,;:!/([{?}])",-1,nc) && nc \|\|
ali@70	2409	!g_utf8_strchr(" _-([{'`",-1,pc) && g_unichar_isalpha(nc))
ali@56	2410	{
ali@56	2411	if (pswit[ECHO_SWITCH])
ali@70	2412	g_print("\n%s\n",aline);
ali@56	2413	if (!pswit[OVERVIEW_SWITCH])
ali@70	2414	g_print(" Line %ld column %ld - Unspaced quotes?\n",
ali@70	2415	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2416	else
ali@56	2417	cnt_punct++;
ali@56	2418	}
ali@56	2419	}
ali@56	2420	}
ali@56	2421	/* Check parity of quotes. */
ali@70	2422	nc=g_utf8_get_char(aline);
ali@70	2423	for (s=aline;*s;s=g_utf8_next_char(s))
ali@56	2424	{
ali@70	2425	c=nc;
ali@70	2426	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@142	2427	if (CHAR_IS_DQUOTE(c))
ali@56	2428	{
ali@142	2429	if (c==CHAR_DQUOTE)
ali@142	2430	{
ali@142	2431	parities->dquote=!parities->dquote;
ali@142	2432	parity=parities->dquote;
ali@142	2433	}
ali@142	2434	else if (c==CHAR_LD_QUOTE)
ali@142	2435	parity=1;
ali@142	2436	else
ali@142	2437	parity=0;
ali@142	2438	if (!parity)
ali@56	2439	{
ali@56	2440	/* parity even */
ali@70	2441	if (!g_utf8_strchr("_-.'`/,;:!?)]} ",-1,nc))
ali@56	2442	{
ali@56	2443	if (pswit[ECHO_SWITCH])
ali@70	2444	g_print("\n%s\n",aline);
ali@56	2445	if (!pswit[OVERVIEW_SWITCH])
ali@70	2446	g_print(" Line %ld column %ld - "
ali@70	2447	"Wrongspaced quotes?\n",
ali@70	2448	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2449	else
ali@56	2450	cnt_punct++;
ali@56	2451	}
ali@56	2452	}
ali@56	2453	else
ali@56	2454	{
ali@56	2455	/* parity odd */
ali@70	2456	if (!g_unichar_isalpha(nc) && !isdigit(nc) &&
ali@70	2457	!g_utf8_strchr("_-/.'`([{$",-1,nc) \|\| !nc)
ali@56	2458	{
ali@56	2459	if (pswit[ECHO_SWITCH])
ali@70	2460	g_print("\n%s\n",aline);
ali@56	2461	if (!pswit[OVERVIEW_SWITCH])
ali@70	2462	g_print(" Line %ld column %ld - "
ali@70	2463	"Wrongspaced quotes?\n",
ali@70	2464	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2465	else
ali@56	2466	cnt_punct++;
ali@56	2467	}
ali@56	2468	}
ali@56	2469	}
ali@56	2470	}
ali@142	2471	c=g_utf8_get_char(aline);
ali@142	2472	if (CHAR_IS_DQUOTE(c))
ali@56	2473	{
ali@70	2474	if (g_utf8_strchr(",;:!?)]} ",-1,
ali@70	2475	g_utf8_get_char(g_utf8_next_char(aline))))
ali@56	2476	{
ali@56	2477	if (pswit[ECHO_SWITCH])
ali@70	2478	g_print("\n%s\n",aline);
ali@56	2479	if (!pswit[OVERVIEW_SWITCH])
ali@70	2480	g_print(" Line %ld column 1 - Wrongspaced quotes?\n",
ali@56	2481	linecnt);
ali@56	2482	else
ali@56	2483	cnt_punct++;
ali@56	2484	}
ali@56	2485	}
ali@56	2486	if (pswit[SQUOTE_SWITCH])
ali@56	2487	{
ali@70	2488	nc=g_utf8_get_char(aline);
ali@70	2489	for (s=aline;*s;s=g_utf8_next_char(s))
ali@56	2490	{
ali@70	2491	c=nc;
ali@70	2492	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@99	2493	if (CHAR_IS_SQUOTE(c) && (s==aline \|\| s>aline &&
ali@70	2494	!g_unichar_isalpha(g_utf8_get_char(g_utf8_prev_char(s))) \|\|
ali@70	2495	!g_unichar_isalpha(nc)))
ali@56	2496	{
ali@56	2497	parities->squote=!parities->squote;
ali@56	2498	if (!parities->squote)
ali@56	2499	{
ali@56	2500	/* parity even */
ali@70	2501	if (!g_utf8_strchr("_-.'`/\",;:!?)]} ",-1,nc))
ali@56	2502	{
ali@56	2503	if (pswit[ECHO_SWITCH])
ali@70	2504	g_print("\n%s\n",aline);
ali@56	2505	if (!pswit[OVERVIEW_SWITCH])
ali@70	2506	g_print(" Line %ld column %ld - "
ali@56	2507	"Wrongspaced singlequotes?\n",
ali@70	2508	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2509	else
ali@56	2510	cnt_punct++;
ali@56	2511	}
ali@56	2512	}
ali@56	2513	else
ali@56	2514	{
ali@56	2515	/* parity odd */
ali@70	2516	if (!g_unichar_isalpha(nc) && !isdigit(nc) &&
ali@70	2517	!g_utf8_strchr("_-/\".'`",-1,nc) \|\| !nc)
ali@56	2518	{
ali@56	2519	if (pswit[ECHO_SWITCH])
ali@70	2520	g_print("\n%s\n",aline);
ali@56	2521	if (!pswit[OVERVIEW_SWITCH])
ali@70	2522	g_print(" Line %ld column %ld - "
ali@56	2523	"Wrongspaced singlequotes?\n",
ali@70	2524	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2525	else
ali@56	2526	cnt_punct++;
ali@56	2527	}
ali@56	2528	}
ali@56	2529	}
ali@56	2530	}
ali@56	2531	}
ali@56	2532	}
ali@56	2533
ali@55	2534	/*
ali@57	2535	* check_for_double_punctuation:
ali@57	2536	*
ali@57	2537	* Look for double punctuation like ,. or ,,
ali@57	2538	* Thanks to DW for the suggestion!
ali@57	2539	* In books with references, ".," and ".;" are common
ali@57	2540	* e.g. "etc., etc.," and vol. 1.; vol 3.;
ali@57	2541	* OTOH, from my initial tests, there are also fairly
ali@57	2542	* common errors. What to do? Make these cases paranoid?
ali@57	2543	* ".," is the most common, so warnings->dotcomma is used
ali@57	2544	* to suppress detailed reporting if it occurs often.
ali@57	2545	*/
ali@57	2546	void check_for_double_punctuation(const char aline,struct warnings warnings)
ali@57	2547	{
ali@70	2548	const char *s;
ali@70	2549	gunichar c,nc;
ali@70	2550	nc=g_utf8_get_char(aline);
ali@70	2551	for (s=aline;*s;s=g_utf8_next_char(s))
ali@57	2552	{
ali@70	2553	c=nc;
ali@70	2554	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@57	2555	/* for each punctuation character in the line */
ali@70	2556	if (c && nc && g_utf8_strchr(".?!,;:",-1,c) &&
ali@70	2557	g_utf8_strchr(".?!,;:",-1,nc))
ali@57	2558	{
ali@57	2559	/* followed by punctuation, it's a query, unless . . . */
ali@70	2560	if (c==nc && (c=='.' \|\| c=='?' \|\| c=='!') \|\|
ali@70	2561	!warnings->dotcomma && c=='.' && nc==',' \|\|
ali@70	2562	warnings->isFrench && g_str_has_prefix(s,",...") \|\|
ali@70	2563	warnings->isFrench && g_str_has_prefix(s,"...,") \|\|
ali@70	2564	warnings->isFrench && g_str_has_prefix(s,";...") \|\|
ali@70	2565	warnings->isFrench && g_str_has_prefix(s,"...;") \|\|
ali@70	2566	warnings->isFrench && g_str_has_prefix(s,":...") \|\|
ali@70	2567	warnings->isFrench && g_str_has_prefix(s,"...:") \|\|
ali@70	2568	warnings->isFrench && g_str_has_prefix(s,"!...") \|\|
ali@70	2569	warnings->isFrench && g_str_has_prefix(s,"...!") \|\|
ali@70	2570	warnings->isFrench && g_str_has_prefix(s,"?...") \|\|
ali@70	2571	warnings->isFrench && g_str_has_prefix(s,"...?"))
ali@57	2572	{
ali@70	2573	if (warnings->isFrench && g_str_has_prefix(s,",...") \|\|
ali@70	2574	warnings->isFrench && g_str_has_prefix(s,"...,") \|\|
ali@70	2575	warnings->isFrench && g_str_has_prefix(s,";...") \|\|
ali@70	2576	warnings->isFrench && g_str_has_prefix(s,"...;") \|\|
ali@70	2577	warnings->isFrench && g_str_has_prefix(s,":...") \|\|
ali@70	2578	warnings->isFrench && g_str_has_prefix(s,"...:") \|\|
ali@70	2579	warnings->isFrench && g_str_has_prefix(s,"!...") \|\|
ali@70	2580	warnings->isFrench && g_str_has_prefix(s,"...!") \|\|
ali@70	2581	warnings->isFrench && g_str_has_prefix(s,"?...") \|\|
ali@70	2582	warnings->isFrench && g_str_has_prefix(s,"...?"))
ali@70	2583	{
ali@70	2584	s+=4;
ali@70	2585	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@70	2586	}
ali@57	2587	; /* do nothing for .. !! and ?? which can be legit */
ali@57	2588	}
ali@57	2589	else
ali@57	2590	{
ali@57	2591	if (pswit[ECHO_SWITCH])
ali@70	2592	g_print("\n%s\n",aline);
ali@57	2593	if (!pswit[OVERVIEW_SWITCH])
ali@70	2594	g_print(" Line %ld column %ld - Double punctuation?\n",
ali@70	2595	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@57	2596	else
ali@57	2597	cnt_punct++;
ali@57	2598	}
ali@57	2599	}
ali@57	2600	}
ali@57	2601	}
ali@57	2602
ali@57	2603	/*
ali@58	2604	* check_for_spaced_quotes:
ali@58	2605	*/
ali@58	2606	void check_for_spaced_quotes(const char *aline)
ali@58	2607	{
ali@99	2608	int i;
ali@58	2609	const char s,t;
ali@99	2610	const gunichar single_quotes[]={CHAR_SQUOTE,CHAR_OPEN_SQUOTE,CHAR_LS_QUOTE,
ali@99	2611	CHAR_RS_QUOTE};
ali@99	2612	GString *pattern;
ali@58	2613	s=aline;
ali@58	2614	while ((t=strstr(s," \" ")))
ali@58	2615	{
ali@58	2616	if (pswit[ECHO_SWITCH])
ali@70	2617	g_print("\n%s\n",aline);
ali@58	2618	if (!pswit[OVERVIEW_SWITCH])
ali@70	2619	g_print(" Line %ld column %ld - Spaced doublequote?\n",
ali@70	2620	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@58	2621	else
ali@58	2622	cnt_punct++;
ali@70	2623	s=g_utf8_next_char(g_utf8_next_char(t));
ali@58	2624	}
ali@99	2625	pattern=g_string_new(NULL);
ali@99	2626	for(i=0;i<G_N_ELEMENTS(single_quotes);i++)
ali@58	2627	{
ali@99	2628	g_string_assign(pattern," ");
ali@99	2629	g_string_append_unichar(pattern,single_quotes[i]);
ali@99	2630	g_string_append_c(pattern,' ');
ali@99	2631	s=aline;
ali@99	2632	while ((t=strstr(s,pattern->str)))
ali@99	2633	{
ali@99	2634	if (pswit[ECHO_SWITCH])
ali@99	2635	g_print("\n%s\n",aline);
ali@99	2636	if (!pswit[OVERVIEW_SWITCH])
ali@99	2637	g_print(" Line %ld column %ld - Spaced singlequote?\n",
ali@99	2638	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@99	2639	else
ali@99	2640	cnt_punct++;
ali@99	2641	s=g_utf8_next_char(g_utf8_next_char(t));
ali@99	2642	}
ali@58	2643	}
ali@99	2644	g_string_free(pattern,TRUE);
ali@58	2645	}
ali@58	2646
ali@58	2647	/*
ali@59	2648	* check_for_miscased_genative:
ali@59	2649	*
ali@59	2650	* Check special case of 'S instead of 's at end of word.
ali@59	2651	*/
ali@59	2652	void check_for_miscased_genative(const char *aline)
ali@59	2653	{
ali@59	2654	const char *s;
ali@70	2655	gunichar c,nc,pc;
ali@69	2656	if (!*aline)
ali@69	2657	return;
ali@70	2658	c=g_utf8_get_char(aline);
ali@70	2659	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2660	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@59	2661	{
ali@70	2662	pc=c;
ali@70	2663	c=nc;
ali@70	2664	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@99	2665	if (CHAR_IS_APOSTROPHE(c) && nc=='S' && g_unichar_islower(pc))
ali@59	2666	{
ali@59	2667	if (pswit[ECHO_SWITCH])
ali@70	2668	g_print("\n%s\n",aline);
ali@59	2669	if (!pswit[OVERVIEW_SWITCH])
ali@70	2670	g_print(" Line %ld column %ld - Capital \"S\"?\n",
ali@70	2671	linecnt,g_utf8_pointer_to_offset(aline,s)+2);
ali@59	2672	else
ali@59	2673	cnt_punct++;
ali@59	2674	}
ali@59	2675	}
ali@59	2676	}
ali@59	2677
ali@59	2678	/*
ali@60	2679	* check_end_of_line:
ali@60	2680	*
ali@60	2681	* Now check special cases - start and end of line -
ali@60	2682	* for single and double quotes. Start is sometimes [sic]
ali@60	2683	* but better to query it anyway.
ali@60	2684	* While we're here, check for dash at end of line.
ali@60	2685	*/
ali@60	2686	void check_end_of_line(const char aline,struct warnings warnings)
ali@60	2687	{
ali@70	2688	int lbytes;
ali@70	2689	const char *s;
ali@70	2690	gunichar c1,c2;
ali@70	2691	lbytes=strlen(aline);
ali@70	2692	if (g_utf8_strlen(aline,lbytes)>1)
ali@60	2693	{
ali@70	2694	s=g_utf8_prev_char(aline+lbytes);
ali@70	2695	c1=g_utf8_get_char(s);
ali@70	2696	c2=g_utf8_get_char(g_utf8_prev_char(s));
ali@142	2697	if ((CHAR_IS_DQUOTE(c1) \|\| CHAR_IS_SQUOTE(c1)) && c2==CHAR_SPACE)
ali@60	2698	{
ali@60	2699	if (pswit[ECHO_SWITCH])
ali@70	2700	g_print("\n%s\n",aline);
ali@60	2701	if (!pswit[OVERVIEW_SWITCH])
ali@70	2702	g_print(" Line %ld column %ld - Spaced quote?\n",linecnt,
ali@70	2703	g_utf8_strlen(aline,lbytes));
ali@70	2704	else
ali@70	2705	cnt_punct++;
ali@70	2706	}
ali@70	2707	c1=g_utf8_get_char(aline);
ali@70	2708	c2=g_utf8_get_char(g_utf8_next_char(aline));
ali@99	2709	if (CHAR_IS_SQUOTE(c1) && c2==CHAR_SPACE)
ali@70	2710	{
ali@70	2711	if (pswit[ECHO_SWITCH])
ali@70	2712	g_print("\n%s\n",aline);
ali@70	2713	if (!pswit[OVERVIEW_SWITCH])
ali@70	2714	g_print(" Line %ld column 1 - Spaced quote?\n",linecnt);
ali@60	2715	else
ali@60	2716	cnt_punct++;
ali@60	2717	}
ali@60	2718	/*
ali@60	2719	* Dash at end of line may well be legit - paranoid mode only
ali@60	2720	* and don't report em-dash at line-end.
ali@60	2721	*/
ali@60	2722	if (pswit[PARANOID_SWITCH] && warnings->hyphen)
ali@60	2723	{
ali@70	2724	for (s=g_utf8_prev_char(aline+lbytes);
ali@70	2725	s>aline && g_utf8_get_char(s)<=CHAR_SPACE;s=g_utf8_prev_char(s))
ali@60	2726	;
ali@70	2727	if (g_utf8_get_char(s)=='-' &&
ali@70	2728	g_utf8_get_char(g_utf8_prev_char(s))!='-')
ali@60	2729	{
ali@60	2730	if (pswit[ECHO_SWITCH])
ali@70	2731	g_print("\n%s\n",aline);
ali@60	2732	if (!pswit[OVERVIEW_SWITCH])
ali@70	2733	g_print(" Line %ld column %ld - "
ali@70	2734	"Hyphen at end of line?\n",
ali@70	2735	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@60	2736	}
ali@60	2737	}
ali@60	2738	}
ali@60	2739	}
ali@60	2740
ali@60	2741	/*
ali@61	2742	* check_for_unspaced_bracket:
ali@61	2743	*
ali@61	2744	* Brackets are often unspaced, but shouldn't be surrounded by alpha.
ali@61	2745	* If so, suspect a scanno like "a]most".
ali@61	2746	*/
ali@61	2747	void check_for_unspaced_bracket(const char *aline)
ali@61	2748	{
ali@70	2749	const char *s;
ali@70	2750	gunichar c,nc,pc;
ali@70	2751	c=g_utf8_get_char(aline);
ali@70	2752	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2753	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@61	2754	{
ali@70	2755	pc=c;
ali@70	2756	c=nc;
ali@70	2757	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@70	2758	if (!nc)
ali@70	2759	break;
ali@61	2760	/* for each bracket character in the line except 1st & last */
ali@70	2761	if (g_utf8_strchr("{[()]}",-1,c) &&
ali@70	2762	g_unichar_isalpha(pc) && g_unichar_isalpha(nc))
ali@61	2763	{
ali@61	2764	if (pswit[ECHO_SWITCH])
ali@70	2765	g_print("\n%s\n",aline);
ali@61	2766	if (!pswit[OVERVIEW_SWITCH])
ali@70	2767	g_print(" Line %ld column %ld - Unspaced bracket?\n",
ali@70	2768	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@61	2769	else
ali@61	2770	cnt_punct++;
ali@61	2771	}
ali@61	2772	}
ali@61	2773	}
ali@61	2774
ali@61	2775	/*
ali@62	2776	* check_for_unpunctuated_endquote:
ali@62	2777	*/
ali@62	2778	void check_for_unpunctuated_endquote(const char *aline)
ali@62	2779	{
ali@70	2780	const char *s;
ali@70	2781	gunichar c,nc,pc;
ali@142	2782	QuoteClass qc;
ali@70	2783	c=g_utf8_get_char(aline);
ali@70	2784	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2785	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@62	2786	{
ali@70	2787	pc=c;
ali@70	2788	c=nc;
ali@142	2789	qc=CHAR_IS_DQUOTE(c)?QUOTE_CLASS(c):INVALID_QUOTE;
ali@70	2790	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@62	2791	/* for each character in the line except 1st */
ali@147	2792	if ((qc==CLOSING_QUOTE \|\| qc==NEUTRAL_QUOTE) && g_unichar_isalpha(pc))
ali@62	2793	{
ali@62	2794	if (pswit[ECHO_SWITCH])
ali@70	2795	g_print("\n%s\n",aline);
ali@62	2796	if (!pswit[OVERVIEW_SWITCH])
ali@70	2797	g_print(" Line %ld column %ld - "
ali@70	2798	"endquote missing punctuation?\n",
ali@70	2799	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@62	2800	else
ali@62	2801	cnt_punct++;
ali@62	2802	}
ali@62	2803	}
ali@62	2804	}
ali@62	2805
ali@62	2806	/*
ali@63	2807	* check_for_html_tag:
ali@63	2808	*
ali@63	2809	* Check for <HTML TAG>.
ali@63	2810	*
ali@63	2811	* If there is a < in the line, followed at some point
ali@63	2812	* by a > then we suspect HTML.
ali@63	2813	*/
ali@63	2814	void check_for_html_tag(const char *aline)
ali@63	2815	{
ali@63	2816	const char open,close;
ali@70	2817	gchar *tag;
ali@70	2818	open=strchr(aline,'<');
ali@63	2819	if (open)
ali@63	2820	{
ali@70	2821	close=strchr(g_utf8_next_char(open),'>');
ali@63	2822	if (close)
ali@63	2823	{
ali@70	2824	if (pswit[ECHO_SWITCH])
ali@70	2825	g_print("\n%s\n",aline);
ali@70	2826	if (!pswit[OVERVIEW_SWITCH])
ali@63	2827	{
ali@70	2828	tag=g_strndup(open,close-open+1);
ali@70	2829	g_print(" Line %ld column %ld - HTML Tag? %s \n",
ali@70	2830	linecnt,g_utf8_pointer_to_offset(aline,open)+1,tag);
ali@70	2831	g_free(tag);
ali@63	2832	}
ali@70	2833	else
ali@70	2834	cnt_html++;
ali@63	2835	}
ali@63	2836	}
ali@63	2837	}
ali@63	2838
ali@63	2839	/*
ali@64	2840	* check_for_html_entity:
ali@64	2841	*
ali@64	2842	* Check for &symbol; HTML.
ali@64	2843	*
ali@64	2844	* If there is a & in the line, followed at
ali@64	2845	* some point by a ; then we suspect HTML.
ali@64	2846	*/
ali@64	2847	void check_for_html_entity(const char *aline)
ali@64	2848	{
ali@64	2849	const char s,amp,*scolon;
ali@70	2850	gchar *entity;
ali@70	2851	amp=strchr(aline,'&');
ali@64	2852	if (amp)
ali@64	2853	{
ali@70	2854	scolon=strchr(amp,';');
ali@64	2855	if (scolon)
ali@64	2856	{
ali@70	2857	for (s=amp;s<scolon;s=g_utf8_next_char(s))
ali@70	2858	if (g_utf8_get_char(s)==CHAR_SPACE)
ali@70	2859	break; /* Don't report "Jones & Son;" */
ali@70	2860	if (s>=scolon)
ali@64	2861	{
ali@64	2862	if (pswit[ECHO_SWITCH])
ali@70	2863	g_print("\n%s\n",aline);
ali@64	2864	if (!pswit[OVERVIEW_SWITCH])
ali@70	2865	{
ali@70	2866	entity=g_strndup(amp,scolon-amp+1);
ali@70	2867	g_print(" Line %ld column %d - HTML symbol? %s \n",
ali@70	2868	linecnt,(int)(amp-aline)+1,entity);
ali@70	2869	g_free(entity);
ali@70	2870	}
ali@64	2871	else
ali@64	2872	cnt_html++;
ali@64	2873	}
ali@64	2874	}
ali@64	2875	}
ali@64	2876	}
ali@64	2877
ali@65	2878	/*
ali@66	2879	* check_for_omitted_punctuation:
ali@66	2880	*
ali@66	2881	* Check for omitted punctuation at end of paragraph by working back
ali@66	2882	* through prevline. DW.
ali@66	2883	* Need to check this only for "normal" paras.
ali@66	2884	* So what is a "normal" para?
ali@66	2885	* Not normal if one-liner (chapter headings, etc.)
ali@66	2886	* Not normal if doesn't contain at least one locase letter
ali@66	2887	* Not normal if starts with space
ali@66	2888	*/
ali@66	2889	void check_for_omitted_punctuation(const char *prevline,
ali@66	2890	struct line_properties *last,int start_para_line)
ali@66	2891	{
ali@70	2892	gboolean letter_on_line=FALSE;
ali@66	2893	const char *s;
ali@99	2894	gunichar c;
ali@142	2895	gboolean closing_quote;
ali@70	2896	for (s=prevline;*s;s=g_utf8_next_char(s))
ali@70	2897	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@70	2898	{
ali@70	2899	letter_on_line=TRUE;
ali@70	2900	break;
ali@70	2901	}
ali@66	2902	/*
ali@66	2903	* This next "if" is a problem.
ali@66	2904	* If we say "start_para_line <= linecnt - 1", that includes
ali@66	2905	* one-line "paragraphs" like chapter heads. Lotsa false positives.
ali@66	2906	* If we say "start_para_line < linecnt - 1" it doesn't, but then it
ali@66	2907	* misses genuine one-line paragraphs.
ali@66	2908	*/
ali@70	2909	if (letter_on_line && last->blen>2 && start_para_line<linecnt-1 &&
ali@70	2910	g_utf8_get_char(prevline)>CHAR_SPACE)
ali@66	2911	{
ali@99	2912	s=prevline+strlen(prevline);
ali@99	2913	do
ali@99	2914	{
ali@99	2915	s=g_utf8_prev_char(s);
ali@99	2916	c=g_utf8_get_char(s);
ali@142	2917	if (QUOTE_CLASS(c)==CLOSING_QUOTE \|\| QUOTE_CLASS(c)==NEUTRAL_QUOTE)
ali@142	2918	closing_quote=TRUE;
ali@142	2919	else
ali@142	2920	closing_quote=FALSE;
ali@142	2921	} while (closing_quote && s>prevline);
ali@70	2922	for (;s>prevline;s=g_utf8_prev_char(s))
ali@66	2923	{
ali@70	2924	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@66	2925	{
ali@66	2926	if (pswit[ECHO_SWITCH])
ali@70	2927	g_print("\n%s\n",prevline);
ali@66	2928	if (!pswit[OVERVIEW_SWITCH])
ali@70	2929	g_print(" Line %ld column %ld - "
ali@66	2930	"No punctuation at para end?\n",
ali@70	2931	linecnt-1,g_utf8_strlen(prevline,-1));
ali@66	2932	else
ali@66	2933	cnt_punct++;
ali@66	2934	break;
ali@66	2935	}
ali@147	2936	if (g_utf8_strchr("—-.:!([{?}])",-1,g_utf8_get_char(s)))
ali@66	2937	break;
ali@66	2938	}
ali@66	2939	}
ali@66	2940	}
ali@66	2941
ali@69	2942	gboolean report_duplicate_queries(gpointer key,gpointer value,gpointer data)
ali@69	2943	{
ali@69	2944	const char *word=key;
ali@69	2945	int *dupcnt=value;
ali@69	2946	if (*dupcnt)
ali@70	2947	g_print("\nNote: Queried word %s was duplicated %d times\n",
ali@69	2948	word,*dupcnt);
ali@69	2949	return FALSE;
ali@69	2950	}
ali@69	2951
ali@70	2952	void print_as_windows_1252(const char *string)
ali@70	2953	{
ali@70	2954	gsize inbytes,outbytes;
ali@70	2955	gchar buf,bp;
ali@86	2956	static GIConv converter=(GIConv)-1;
ali@70	2957	if (!string)
ali@70	2958	{
ali@70	2959	if (converter!=(GIConv)-1)
ali@70	2960	g_iconv_close(converter);
ali@70	2961	converter=(GIConv)-1;
ali@70	2962	return;
ali@70	2963	}
ali@86	2964	if (converter==(GIConv)-1)
ali@70	2965	converter=g_iconv_open("WINDOWS-1252","UTF-8");
ali@70	2966	if (converter!=(GIConv)-1)
ali@70	2967	{
ali@70	2968	inbytes=outbytes=strlen(string);
ali@70	2969	bp=buf=g_malloc(outbytes+1);
ali@70	2970	g_iconv(converter,(char **)&string,&inbytes,&bp,&outbytes);
ali@70	2971	*bp='\0';
ali@70	2972	fputs(buf,stdout);
ali@70	2973	g_free(buf);
ali@70	2974	}
ali@70	2975	else
ali@70	2976	fputs(string,stdout);
ali@70	2977	}
ali@70	2978
ali@72	2979	void print_as_utf_8(const char *string)
ali@72	2980	{
ali@72	2981	fputs(string,stdout);
ali@72	2982	}
ali@72	2983
ali@66	2984	/*
ali@41	2985	* procfile:
ali@41	2986	*
ali@41	2987	* Process one file.
ali@41	2988	*/
ali@69	2989	void procfile(const char *filename)
ali@41	2990	{
ali@65	2991	const char *s;
ali@69	2992	gchar parastart=NULL; / first line of current para */
ali@69	2993	gchar etext,aline;
ali@69	2994	gchar *etext_ptr;
ali@69	2995	GError *err=NULL;
ali@41	2996	struct first_pass_results *first_pass_results;
ali@42	2997	struct warnings *warnings;
ali@43	2998	struct counters counters={0};
ali@45	2999	struct line_properties last={0};
ali@56	3000	struct parities parities={0};
ali@69	3001	struct pending pending={0};
ali@69	3002	gboolean isemptyline;
ali@68	3003	long start_para_line=0;
ali@69	3004	gboolean isnewpara=FALSE,enddash=FALSE;
ali@45	3005	last.start=CHAR_SPACE;
ali@68	3006	linecnt=checked_linecnt=0;
ali@69	3007	etext=read_etext(filename,&err);
ali@69	3008	if (!etext)
ali@41	3009	{
ali@68	3010	if (pswit[STDOUT_SWITCH])
ali@69	3011	fprintf(stdout,"bookloupe: %s: %s\n",filename,err->message);
ali@68	3012	else
ali@69	3013	fprintf(stderr,"bookloupe: %s: %s\n",filename,err->message);
ali@41	3014	exit(1);
ali@41	3015	}
ali@70	3016	g_print("\n\nFile: %s\n\n",filename);
ali@69	3017	first_pass_results=first_pass(etext);
ali@42	3018	warnings=report_first_pass(first_pass_results);
ali@69	3019	qword=g_tree_new_full((GCompareDataFunc)strcmp,NULL,g_free,g_free);
ali@69	3020	qperiod=g_tree_new_full((GCompareDataFunc)strcmp,NULL,g_free,NULL);
ali@40	3021	/*
ali@40	3022	* Here we go with the main pass. Hold onto yer hat!
ali@40	3023	*/
ali@65	3024	linecnt=0;
ali@69	3025	etext_ptr=etext;
ali@69	3026	while ((aline=flgets(&etext_ptr,linecnt+1)))
ali@40	3027	{
ali@68	3028	linecnt++;
ali@68	3029	if (linecnt==1)
ali@69	3030	isnewpara=TRUE;
ali@70	3031	if (pswit[DP_SWITCH] && g_str_has_prefix(aline,"-----File: "))
ali@40	3032	continue; // skip DP page separators completely
ali@68	3033	if (linecnt<first_pass_results->firstline \|\|
ali@41	3034	(first_pass_results->footerline>0 &&
ali@41	3035	linecnt>first_pass_results->footerline))
ali@40	3036	{
ali@68	3037	if (pswit[HEADER_SWITCH])
ali@40	3038	{
ali@70	3039	if (g_str_has_prefix(aline,"Title:"))
ali@70	3040	g_print(" %s\n",aline);
ali@70	3041	if (g_str_has_prefix(aline,"Author:"))
ali@70	3042	g_print(" %s\n",aline);
ali@70	3043	if (g_str_has_prefix(aline,"Release Date:"))
ali@70	3044	g_print(" %s\n",aline);
ali@70	3045	if (g_str_has_prefix(aline,"Edition:"))
ali@70	3046	g_print(" %s\n\n",aline);
ali@40	3047	}
ali@68	3048	continue; /* skip through the header */
ali@40	3049	}
ali@68	3050	checked_linecnt++;
ali@65	3051	print_pending(aline,parastart,&pending);
ali@164	3052	isemptyline=analyse_quotes(aline,&counters);
ali@68	3053	if (isnewpara && !isemptyline)
ali@40	3054	{
ali@40	3055	/* This line is the start of a new paragraph. */
ali@68	3056	start_para_line=linecnt;
ali@40	3057	/* Capture its first line in case we want to report it later. */
ali@69	3058	g_free(parastart);
ali@69	3059	parastart=g_strdup(aline);
ali@56	3060	memset(&parities,0,sizeof(parities)); /* restart the quote count */
ali@68	3061	s=aline;
ali@70	3062	while (*s && !g_unichar_isalpha(g_utf8_get_char(s)) &&
ali@70	3063	!g_unichar_isdigit(g_utf8_get_char(s)))
ali@70	3064	s=g_utf8_next_char(s);
ali@70	3065	if (g_unichar_islower(g_utf8_get_char(s)))
ali@40	3066	{
ali@40	3067	/* and its first letter is lowercase */
ali@68	3068	if (pswit[ECHO_SWITCH])
ali@70	3069	g_print("\n%s\n",aline);
ali@68	3070	if (!pswit[OVERVIEW_SWITCH])
ali@70	3071	g_print(" Line %ld column %ld - "
ali@40	3072	"Paragraph starts with lower-case\n",
ali@70	3073	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@68	3074	else
ali@68	3075	cnt_punct++;
ali@40	3076	}
ali@69	3077	isnewpara=FALSE; /* Signal the end of new para processing. */
ali@40	3078	}
ali@68	3079	/* Check for an em-dash broken at line end. */
ali@70	3080	if (enddash && g_utf8_get_char(aline)=='-')
ali@40	3081	{
ali@68	3082	if (pswit[ECHO_SWITCH])
ali@70	3083	g_print("\n%s\n",aline);
ali@68	3084	if (!pswit[OVERVIEW_SWITCH])
ali@70	3085	g_print(" Line %ld column 1 - Broken em-dash?\n",linecnt);
ali@68	3086	else
ali@68	3087	cnt_punct++;
ali@40	3088	}
ali@69	3089	enddash=FALSE;
ali@70	3090	for (s=g_utf8_prev_char(aline+strlen(aline));
ali@70	3091	g_utf8_get_char(s)==' ' && s>aline;s=g_utf8_prev_char(s))
ali@40	3092	;
ali@70	3093	if (s>=aline && g_utf8_get_char(s)=='-')
ali@69	3094	enddash=TRUE;
ali@67	3095	check_for_control_characters(aline);
ali@166	3096	check_for_odd_characters(aline,warnings,isemptyline);
ali@68	3097	if (warnings->longline)
ali@45	3098	check_for_long_line(aline);
ali@68	3099	if (warnings->shortline)
ali@45	3100	check_for_short_line(aline,&last);
ali@68	3101	last.blen=last.len;
ali@70	3102	last.len=g_utf8_strlen(aline,-1);
ali@70	3103	last.start=g_utf8_get_char(aline);
ali@46	3104	check_for_starting_punctuation(aline);
ali@68	3105	if (warnings->dash)
ali@40	3106	{
ali@47	3107	check_for_spaced_emdash(aline);
ali@47	3108	check_for_spaced_dash(aline);
ali@40	3109	}
ali@48	3110	check_for_unmarked_paragraphs(aline);
ali@49	3111	check_for_jeebies(aline);
ali@50	3112	check_for_mta_from(aline);
ali@51	3113	check_for_orphan_character(aline);
ali@52	3114	check_for_pling_scanno(aline);
ali@53	3115	check_for_extra_period(aline,warnings);
ali@54	3116	check_for_following_punctuation(aline);
ali@55	3117	check_for_typos(aline,warnings);
ali@56	3118	check_for_misspaced_punctuation(aline,&parities,isemptyline);
ali@57	3119	check_for_double_punctuation(aline,warnings);
ali@58	3120	check_for_spaced_quotes(aline);
ali@59	3121	check_for_miscased_genative(aline);
ali@60	3122	check_end_of_line(aline,warnings);
ali@61	3123	check_for_unspaced_bracket(aline);
ali@68	3124	if (warnings->endquote)
ali@62	3125	check_for_unpunctuated_endquote(aline);
ali@63	3126	check_for_html_tag(aline);
ali@64	3127	check_for_html_entity(aline);
ali@68	3128	if (isemptyline)
ali@40	3129	{
ali@65	3130	check_for_mismatched_quotes(&counters,&pending);
ali@103	3131	counters_reset(&counters);
ali@40	3132	/* let the next iteration know that it's starting a new para */
ali@69	3133	isnewpara=TRUE;
ali@69	3134	if (prevline)
ali@69	3135	check_for_omitted_punctuation(prevline,&last,start_para_line);
ali@40	3136	}
ali@69	3137	g_free(prevline);
ali@69	3138	prevline=g_strdup(aline);
ali@0	3139	}
ali@103	3140	linecnt++;
ali@103	3141	check_for_mismatched_quotes(&counters,&pending);
ali@103	3142	print_pending(NULL,parastart,&pending);
ali@103	3143	reset_pending(&pending);
ali@69	3144	if (prevline)
ali@69	3145	{
ali@69	3146	g_free(prevline);
ali@69	3147	prevline=NULL;
ali@69	3148	}
ali@69	3149	g_free(parastart);
ali@69	3150	g_free(prevline);
ali@69	3151	g_free(etext);
ali@79	3152	if (!pswit[OVERVIEW_SWITCH] && !pswit[VERBOSE_SWITCH])
ali@69	3153	g_tree_foreach(qword,report_duplicate_queries,NULL);
ali@69	3154	g_tree_unref(qword);
ali@69	3155	g_tree_unref(qperiod);
ali@99	3156	counters_destroy(&counters);
ali@70	3157	g_set_print_handler(NULL);
ali@70	3158	print_as_windows_1252(NULL);
ali@71	3159	if (pswit[MARKUP_SWITCH])
ali@71	3160	loseentities(NULL);
ali@0	3161	}
ali@0	3162
ali@40	3163	/*
ali@40	3164	* flgets:
ali@40	3165	*
ali@69	3166	* Get one line from the input text, checking for
ali@40	3167	* the existence of exactly one CR/LF line-end per line.
ali@40	3168	*
ali@40	3169	* Returns: a pointer to the line.
ali@40	3170	*/
ali@69	3171	char flgets(char *etext,long lcnt)
ali@0	3172	{
ali@70	3173	gunichar c;
ali@69	3174	gboolean isCR=FALSE;
ali@69	3175	char theline=etext;
ali@70	3176	char *eos=theline;
ali@70	3177	gchar *s;
ali@70	3178	for (;;)
ali@40	3179	{
ali@70	3180	c=g_utf8_get_char(*etext);
ali@70	3181	etext=g_utf8_next_char(etext);
ali@69	3182	if (!c)
ali@68	3183	return NULL;
ali@40	3184	/* either way, it's end of line */
ali@69	3185	if (c=='\n')
ali@40	3186	{
ali@68	3187	if (isCR)
ali@68	3188	break;
ali@68	3189	else
ali@40	3190	{
ali@40	3191	/* Error - a LF without a preceding CR */
ali@68	3192	if (pswit[LINE_END_SWITCH])
ali@40	3193	{
ali@68	3194	if (pswit[ECHO_SWITCH])
ali@70	3195	{
ali@70	3196	s=g_strndup(theline,eos-theline);
ali@70	3197	g_print("\n%s\n",s);
ali@70	3198	g_free(s);
ali@70	3199	}
ali@68	3200	if (!pswit[OVERVIEW_SWITCH])
ali@70	3201	g_print(" Line %ld - No CR?\n",lcnt);
ali@68	3202	else
ali@68	3203	cnt_lineend++;
ali@40	3204	}
ali@68	3205	break;
ali@40	3206	}
ali@40	3207	}
ali@69	3208	if (c=='\r')
ali@40	3209	{
ali@68	3210	if (isCR)
ali@40	3211	{
ali@40	3212	/* Error - two successive CRs */
ali@68	3213	if (pswit[LINE_END_SWITCH])
ali@40	3214	{
ali@68	3215	if (pswit[ECHO_SWITCH])
ali@70	3216	{
ali@70	3217	s=g_strndup(theline,eos-theline);
ali@70	3218	g_print("\n%s\n",s);
ali@70	3219	g_free(s);
ali@70	3220	}
ali@68	3221	if (!pswit[OVERVIEW_SWITCH])
ali@70	3222	g_print(" Line %ld - Two successive CRs?\n",lcnt);
ali@68	3223	else
ali@68	3224	cnt_lineend++;
ali@40	3225	}
ali@40	3226	}
ali@69	3227	isCR=TRUE;
ali@40	3228	}
ali@68	3229	else
ali@40	3230	{
ali@68	3231	if (pswit[LINE_END_SWITCH] && isCR)
ali@40	3232	{
ali@68	3233	if (pswit[ECHO_SWITCH])
ali@70	3234	{
ali@70	3235	s=g_strndup(theline,eos-theline);
ali@70	3236	g_print("\n%s\n",s);
ali@70	3237	g_free(s);
ali@70	3238	}
ali@68	3239	if (!pswit[OVERVIEW_SWITCH])
ali@70	3240	g_print(" Line %ld column %ld - CR without LF?\n",
ali@70	3241	lcnt,g_utf8_pointer_to_offset(theline,eos)+1);
ali@68	3242	else
ali@68	3243	cnt_lineend++;
ali@70	3244	*eos=' ';
ali@40	3245	}
ali@69	3246	isCR=FALSE;
ali@70	3247	eos=g_utf8_next_char(eos);
ali@40	3248	}
ali@69	3249	}
ali@70	3250	*eos='\0';
ali@0	3251	if (pswit[MARKUP_SWITCH])
ali@68	3252	postprocess_for_HTML(theline);
ali@0	3253	if (pswit[DP_SWITCH])
ali@68	3254	postprocess_for_DP(theline);
ali@40	3255	return theline;
ali@0	3256	}
ali@0	3257
ali@40	3258	/*
ali@40	3259	* mixdigit:
ali@40	3260	*
ali@40	3261	* Takes a "word" as a parameter, and checks whether it
ali@40	3262	* contains a mixture of alpha and digits. Generally, this is an
ali@40	3263	* error, but may not be for cases like 4th or L5 12s. 3d.
ali@40	3264	*
ali@70	3265	* Returns: TRUE iff an is error found.
ali@40	3266	*/
ali@70	3267	gboolean mixdigit(const char *checkword)
ali@0	3268	{
ali@70	3269	gboolean wehaveadigit,wehavealetter,query;
ali@70	3270	const char s,nondigit;
ali@70	3271	wehaveadigit=wehavealetter=query=FALSE;
ali@70	3272	for (s=checkword;*s;s=g_utf8_next_char(s))
ali@70	3273	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@70	3274	wehavealetter=TRUE;
ali@70	3275	else if (g_unichar_isdigit(g_utf8_get_char(s)))
ali@70	3276	wehaveadigit=TRUE;
ali@40	3277	if (wehaveadigit && wehavealetter)
ali@40	3278	{
ali@40	3279	/* Now exclude common legit cases, like "21st" and "12l. 3s. 11d." */
ali@70	3280	query=TRUE;
ali@70	3281	for (nondigit=checkword;g_unichar_isdigit(g_utf8_get_char(nondigit));
ali@70	3282	nondigit=g_utf8_next_char(nondigit))
ali@68	3283	;
ali@68	3284	/* digits, ending in st, rd, nd, th of either case */
ali@70	3285	if (!g_ascii_strcasecmp(nondigit,"st") \|\|
ali@70	3286	!g_ascii_strcasecmp(nondigit,"rd") \|\|
ali@70	3287	!g_ascii_strcasecmp(nondigit,"nd") \|\|
ali@70	3288	!g_ascii_strcasecmp(nondigit,"th"))
ali@70	3289	query=FALSE;
ali@70	3290	if (!g_ascii_strcasecmp(nondigit,"sts") \|\|
ali@70	3291	!g_ascii_strcasecmp(nondigit,"rds") \|\|
ali@70	3292	!g_ascii_strcasecmp(nondigit,"nds") \|\|
ali@70	3293	!g_ascii_strcasecmp(nondigit,"ths"))
ali@70	3294	query=FALSE;
ali@70	3295	if (!g_ascii_strcasecmp(nondigit,"stly") \|\|
ali@70	3296	!g_ascii_strcasecmp(nondigit,"rdly") \|\|
ali@70	3297	!g_ascii_strcasecmp(nondigit,"ndly") \|\|
ali@70	3298	!g_ascii_strcasecmp(nondigit,"thly"))
ali@70	3299	query=FALSE;
ali@68	3300	/* digits, ending in l, L, s or d */
ali@70	3301	if (!g_ascii_strcasecmp(nondigit,"l") \|\| !strcmp(nondigit,"s") \|\|
ali@70	3302	!strcmp(nondigit,"d"))
ali@70	3303	query=FALSE;
ali@68	3304	/*
ali@40	3305	* L at the start of a number, representing Britsh pounds, like L500.
ali@70	3306	* This is cute. We know the current word is mixed digit. If the first
ali@68	3307	* letter is L, there must be at least one digit following. If both
ali@68	3308	* digits and letters follow, we have a genuine error, else we have a
ali@68	3309	* capital L followed by digits, and we accept that as a non-error.
ali@40	3310	*/
ali@70	3311	if (g_utf8_get_char(checkword)=='L' &&
ali@70	3312	!mixdigit(g_utf8_next_char(checkword)))
ali@70	3313	query=FALSE;
ali@40	3314	}
ali@40	3315	return query;
ali@0	3316	}
ali@0	3317
ali@40	3318	/*
ali@40	3319	* getaword:
ali@40	3320	*
ali@69	3321	* Extracts the first/next "word" from the line, and returns it.
ali@69	3322	* A word is defined as one English word unit--or at least that's the aim.
ali@69	3323	* "ptr" is advanced to the position in the line where we will start
ali@69	3324	* looking for the next word.
ali@40	3325	*
ali@69	3326	* Returns: A newly-allocated string.
ali@40	3327	*/
ali@69	3328	gchar getaword(const char *ptr)
ali@0	3329	{
ali@70	3330	const char s,t;
ali@69	3331	GString *word;
ali@70	3332	gunichar c,pc;
ali@69	3333	word=g_string_new(NULL);
ali@70	3334	for (;!g_unichar_isdigit(g_utf8_get_char(*ptr)) &&
ali@70	3335	!g_unichar_isalpha(g_utf8_get_char(*ptr)) &&
ali@70	3336	*ptr;ptr=g_utf8_next_char(*ptr))
ali@40	3337	;
ali@40	3338	/*
ali@40	3339	* Use a look-ahead to handle exceptions for numbers like 1,000 and 1.35.
ali@40	3340	* Especially yucky is the case of L1,000
ali@40	3341	* This section looks for a pattern of characters including a digit
ali@40	3342	* followed by a comma or period followed by one or more digits.
ali@40	3343	* If found, it returns this whole pattern as a word; otherwise we discard
ali@40	3344	* the results and resume our normal programming.
ali@40	3345	*/
ali@69	3346	s=*ptr;
ali@70	3347	for (;g_unichar_isdigit(g_utf8_get_char(s)) \|\|
ali@70	3348	g_unichar_isalpha(g_utf8_get_char(s)) \|\|
ali@70	3349	g_utf8_get_char(s)==',' \|\| g_utf8_get_char(s)=='.';s=g_utf8_next_char(s))
ali@70	3350	g_string_append_unichar(word,g_utf8_get_char(s));
ali@82	3351	if (word->len)
ali@40	3352	{
ali@82	3353	for (t=g_utf8_next_char(word->str);*t;t=g_utf8_next_char(t))
ali@40	3354	{
ali@82	3355	c=g_utf8_get_char(t);
ali@82	3356	pc=g_utf8_get_char(g_utf8_prev_char(t));
ali@82	3357	if ((c=='.' \|\| c==',') && g_unichar_isdigit(pc))
ali@82	3358	{
ali@82	3359	*ptr=s;
ali@82	3360	return g_string_free(word,FALSE);
ali@82	3361	}
ali@40	3362	}
ali@40	3363	}
ali@0	3364	/* we didn't find a punctuated number - do the regular getword thing */
ali@69	3365	g_string_truncate(word,0);
ali@99	3366	c=g_utf8_get_char(*ptr);
ali@99	3367	for (;g_unichar_isdigit(c) \|\| g_unichar_isalpha(c) \|\| CHAR_IS_APOSTROPHE(c);
ali@99	3368	ptr=g_utf8_next_char(ptr),c=g_utf8_get_char(*ptr))
ali@99	3369	g_string_append_unichar(word,c);
ali@69	3370	return g_string_free(word,FALSE);
ali@0	3371	}
ali@0	3372
ali@40	3373	/*
ali@40	3374	* isroman:
ali@40	3375	*
ali@40	3376	* Is this word a Roman Numeral?
ali@40	3377	*
ali@40	3378	* It doesn't actually validate that the number is a valid Roman Numeral--for
ali@40	3379	* example it will pass MXXXXXXXXXX as a valid Roman Numeral, but that's not
ali@40	3380	* what we're here to do. If it passes this, it LOOKS like a Roman numeral.
ali@40	3381	* Anyway, the actual Romans were pretty tolerant of bad arithmetic, or
ali@40	3382	* expressions thereof, except when it came to taxes. Allow any number of M,
ali@40	3383	* an optional D, an optional CM or CD, any number of optional Cs, an optional
ali@40	3384	* XL or an optional XC, an optional IX or IV, an optional V and any number
ali@40	3385	* of optional Is.
ali@40	3386	*/
ali@69	3387	gboolean isroman(const char *t)
ali@0	3388	{
ali@69	3389	const char *s;
ali@40	3390	if (!t \|\| !*t)
ali@69	3391	return FALSE;
ali@40	3392	s=t;
ali@70	3393	while (g_utf8_get_char(t)=='m' && *t)
ali@40	3394	t++;
ali@70	3395	if (g_utf8_get_char(t)=='d')
ali@40	3396	t++;
ali@70	3397	if (g_str_has_prefix(t,"cm"))
ali@40	3398	t+=2;
ali@70	3399	if (g_str_has_prefix(t,"cd"))
ali@40	3400	t+=2;
ali@70	3401	while (g_utf8_get_char(t)=='c' && *t)
ali@40	3402	t++;
ali@70	3403	if (g_str_has_prefix(t,"xl"))
ali@40	3404	t+=2;
ali@70	3405	if (g_str_has_prefix(t,"xc"))
ali@40	3406	t+=2;
ali@70	3407	if (g_utf8_get_char(t)=='l')
ali@40	3408	t++;
ali@70	3409	while (g_utf8_get_char(t)=='x' && *t)
ali@40	3410	t++;
ali@70	3411	if (g_str_has_prefix(t,"ix"))
ali@40	3412	t+=2;
ali@70	3413	if (g_str_has_prefix(t,"iv"))
ali@40	3414	t+=2;
ali@70	3415	if (g_utf8_get_char(t)=='v')
ali@40	3416	t++;
ali@70	3417	while (g_utf8_get_char(t)=='i' && *t)
ali@40	3418	t++;
ali@40	3419	return !*t;
ali@0	3420	}
ali@0	3421
ali@40	3422	/*
ali@40	3423	* postprocess_for_DP:
ali@40	3424	*
ali@40	3425	* Invoked with the -d switch from flgets().
ali@40	3426	* It simply "removes" from the line a hard-coded set of common
ali@40	3427	* DP-specific tags, so that the line passed to the main routine has
ali@40	3428	* been pre-cleaned of DP markup.
ali@40	3429	*/
ali@0	3430	void postprocess_for_DP(char *theline)
ali@0	3431	{
ali@40	3432	char s,t;
ali@0	3433	int i;
ali@0	3434	if (!*theline)
ali@68	3435	return;
ali@40	3436	for (i=0;*DPmarkup[i];i++)
ali@70	3437	while ((s=strstr(theline,DPmarkup[i])))
ali@40	3438	{
ali@68	3439	t=s+strlen(DPmarkup[i]);
ali@70	3440	memmove(s,t,strlen(t)+1);
ali@40	3441	}
ali@0	3442	}
ali@0	3443
ali@40	3444	/*
ali@40	3445	* postprocess_for_HTML:
ali@40	3446	*
ali@40	3447	* Invoked with the -m switch from flgets().
ali@40	3448	* It simply "removes" from the line a hard-coded set of common
ali@40	3449	* HTML tags and "replaces" a hard-coded set of common HTML
ali@40	3450	* entities, so that the line passed to the main routine has
ali@40	3451	* been pre-cleaned of HTML.
ali@40	3452	*/
ali@0	3453	void postprocess_for_HTML(char *theline)
ali@0	3454	{
ali@70	3455	while (losemarkup(theline))
ali@70	3456	;
ali@71	3457	loseentities(theline);
ali@0	3458	}
ali@0	3459
ali@0	3460	char losemarkup(char theline)
ali@0	3461	{
ali@40	3462	char s,t;
ali@0	3463	int i;
ali@70	3464	s=strchr(theline,'<');
ali@70	3465	t=s?strchr(s,'>'):NULL;
ali@40	3466	if (!s \|\| !t)
ali@40	3467	return NULL;
ali@40	3468	for (i=0;*markup[i];i++)
ali@70	3469	if (tagcomp(g_utf8_next_char(s),markup[i]))
ali@40	3470	{
ali@70	3471	t=g_utf8_next_char(t);
ali@70	3472	memmove(s,t,strlen(t)+1);
ali@70	3473	return s;
ali@68	3474	}
ali@40	3475	/* It's an unrecognized <xxx>. */
ali@40	3476	return NULL;
ali@0	3477	}
ali@0	3478
ali@71	3479	void loseentities(char *theline)
ali@0	3480	{
ali@0	3481	int i;
ali@71	3482	gsize nb;
ali@71	3483	char amp,scolon;
ali@71	3484	gchar s,t;
ali@71	3485	gunichar c;
ali@71	3486	GTree *entities=NULL;
ali@86	3487	static GIConv translit=(GIConv)-1,to_utf8=(GIConv)-1;
ali@71	3488	if (!theline)
ali@40	3489	{
ali@71	3490	if (entities)
ali@71	3491	g_tree_destroy(entities);
ali@71	3492	entities=NULL;
ali@86	3493	if (translit!=(GIConv)-1)
ali@71	3494	g_iconv_close(translit);
ali@71	3495	translit=(GIConv)-1;
ali@86	3496	if (to_utf8!=(GIConv)-1)
ali@71	3497	g_iconv_close(to_utf8);
ali@71	3498	to_utf8=(GIConv)-1;
ali@71	3499	return;
ali@71	3500	}
ali@71	3501	if (!*theline)
ali@71	3502	return;
ali@71	3503	if (!entities)
ali@71	3504	{
ali@71	3505	entities=g_tree_new((GCompareFunc)strcmp);
ali@71	3506	for(i=0;i<G_N_ELEMENTS(HTMLentities);i++)
ali@71	3507	g_tree_insert(entities,HTMLentities[i].name,
ali@71	3508	GUINT_TO_POINTER(HTMLentities[i].c));
ali@71	3509	}
ali@71	3510	if (translit==(GIConv)-1)
ali@71	3511	translit=g_iconv_open("ISO_8859-1//TRANSLIT","UTF-8");
ali@71	3512	if (to_utf8==(GIConv)-1)
ali@71	3513	to_utf8=g_iconv_open("UTF-8","ISO_8859-1");
ali@71	3514	while((amp=strchr(theline,'&')))
ali@71	3515	{
ali@71	3516	scolon=strchr(amp,';');
ali@71	3517	if (scolon)
ali@40	3518	{
ali@71	3519	if (amp[1]=='#')
ali@71	3520	{
ali@71	3521	if (amp+2+strspn(amp+2,"0123456789")==scolon)
ali@71	3522	c=strtol(amp+2,NULL,10);
ali@71	3523	else if (amp[2]=='x' &&
ali@71	3524	amp+3+strspn(amp+3,"0123456789abcdefABCDEF")==scolon)
ali@71	3525	c=strtol(amp+3,NULL,16);
ali@71	3526	}
ali@71	3527	else
ali@71	3528	{
ali@71	3529	s=g_strndup(amp+1,scolon-(amp+1));
ali@71	3530	c=GPOINTER_TO_UINT(g_tree_lookup(entities,s));
ali@71	3531	g_free(s);
ali@71	3532	}
ali@40	3533	}
ali@71	3534	else
ali@71	3535	c=0;
ali@71	3536	if (c)
ali@71	3537	{
ali@71	3538	theline=amp;
ali@71	3539	if (c<128 \|\| c>=192 && c<=255) /* An ISO-8859-1 character */
ali@71	3540	theline+=g_unichar_to_utf8(c,theline);
ali@71	3541	else
ali@71	3542	{
ali@71	3543	s=g_malloc(6);
ali@71	3544	nb=g_unichar_to_utf8(c,s);
ali@71	3545	t=g_convert_with_iconv(s,nb,translit,NULL,&nb,NULL);
ali@71	3546	g_free(s);
ali@71	3547	s=g_convert_with_iconv(t,nb,to_utf8,NULL,&nb,NULL);
ali@71	3548	g_free(t);
ali@71	3549	memcpy(theline,s,nb);
ali@71	3550	g_free(s);
ali@71	3551	theline+=nb;
ali@71	3552	}
ali@71	3553	memmove(theline,g_utf8_next_char(scolon),
ali@71	3554	strlen(g_utf8_next_char(scolon))+1);
ali@71	3555	}
ali@71	3556	else
ali@71	3557	theline=g_utf8_next_char(amp);
ali@40	3558	}
ali@0	3559	}
ali@0	3560
ali@70	3561	gboolean tagcomp(const char strin,const char basetag)
ali@0	3562	{
ali@70	3563	gboolean retval;
ali@70	3564	gchar s,t;
ali@70	3565	if (g_utf8_get_char(strin)=='/')
ali@70	3566	t=g_utf8_casefold(g_utf8_next_char(strin),-1); /* ignore a slash */
ali@70	3567	else
ali@70	3568	t=g_utf8_casefold(strin,-1);
ali@70	3569	s=g_utf8_casefold(basetag,-1);
ali@70	3570	retval=g_str_has_prefix(t,s);
ali@70	3571	g_free(s);
ali@70	3572	g_free(t);
ali@70	3573	return retval;
ali@0	3574	}
ali@0	3575
ali@69	3576	void proghelp(GOptionContext *context)
ali@0	3577	{
ali@69	3578	gchar *help;
ali@40	3579	fputs("Bookloupe version " PACKAGE_VERSION ".\n",stderr);
ali@40	3580	fputs("Copyright 2000-2005 Jim Tinsley <jtinsley@pobox.com>.\n",stderr);
ali@40	3581	fputs("Copyright 2012- J. Ali Harlow <ali@juiblex.co.uk>.\n",stderr);
ali@40	3582	fputs("Bookloupe comes wih ABSOLUTELY NO WARRANTY. "
ali@40	3583	"For details, read the file COPYING.\n",stderr);
ali@40	3584	fputs("This is Free Software; "
ali@40	3585	"you may redistribute it under certain conditions (GPL);\n",stderr);
ali@40	3586	fputs("read the file COPYING for details.\n\n",stderr);
ali@69	3587	help=g_option_context_get_help(context,TRUE,NULL);
ali@69	3588	fputs(help,stderr);
ali@69	3589	g_free(help);
ali@69	3590	fputs("Sample usage: bookloupe warpeace.txt\n\n",stderr);
ali@40	3591	fputs("Bookloupe queries anything it thinks shouldn't be in a PG text; "
ali@40	3592	"non-ASCII\n",stderr);
ali@40	3593	fputs("characters like accented letters, "
ali@40	3594	"lines longer than 75 or shorter than 55,\n",stderr);
ali@40	3595	fputs("unbalanced quotes or brackets, "
ali@40	3596	"a variety of badly formatted punctuation, \n",stderr);
ali@40	3597	fputs("HTML tags, some likely typos. "
ali@40	3598	"It is NOT a substitute for human judgement.\n",stderr);
ali@0	3599	fputs("\n",stderr);
ali@0	3600	}

author	ali <ali@juiblex.co.uk>
	Wed Oct 16 20:54:58 2013 +0100 (2013-10-16)
changeset 169	d14a95cb0aab
parent 168	ff5126920618
child 170	8169aeaa2790
permissions	-rw-r--r--