bookloupe-testing: bookloupe/bookloupe.c@1aeda7fe17ca (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@185	35	gchar charset; / Or NULL for auto (ISO_8859-1/ASCII or UNICODE) */
ali@185	36	GIConv charset_validator=(GIConv)-1;
ali@185	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@185	133	gchar *opt_charset;
ali@0	134
ali@186	135	gboolean typo_compat,paranoid_compat;
ali@186	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@186	140	{ "no-dp", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	141	G_OPTION_ARG_NONE, pswit+DP_SWITCH,
ali@186	142	"Don't ignore DP-specific markup", NULL },
ali@186	143	{ "echo", 0, G_OPTION_FLAG_HIDDEN, G_OPTION_ARG_NONE, pswit+ECHO_SWITCH,
ali@186	144	"Echo queried line", NULL },
ali@186	145	{ "no-echo", 'e', G_OPTION_FLAG_REVERSE,
ali@186	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@186	150	{ "no-squote", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	151	G_OPTION_ARG_NONE, pswit+SQUOTE_SWITCH,
ali@186	152	"Don't check single quotes", NULL },
ali@186	153	{ "typo", 0, 0, G_OPTION_ARG_NONE, pswit+TYPO_SWITCH,
ali@69	154	"Check common typos", NULL },
ali@186	155	{ "no-typo", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	156	G_OPTION_ARG_NONE, pswit+TYPO_SWITCH,
ali@186	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@186	160	{ "no-qpara", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	161	G_OPTION_ARG_NONE, pswit+QPARA_SWITCH,
ali@186	162	"Don't require closure of quotes on every paragraph", NULL },
ali@186	163	{ "paranoid", 0, G_OPTION_FLAG_HIDDEN,
ali@186	164	G_OPTION_ARG_NONE, pswit+PARANOID_SWITCH,
ali@186	165	"Enable paranoid querying of everything", NULL },
ali@186	166	{ "no-paranoid", 0, G_OPTION_FLAG_REVERSE,
ali@186	167	G_OPTION_ARG_NONE, pswit+PARANOID_SWITCH,
ali@69	168	"Disable paranoid querying of everything", NULL },
ali@186	169	{ "line-end", 0, G_OPTION_FLAG_HIDDEN,
ali@186	170	G_OPTION_ARG_NONE, pswit+LINE_END_SWITCH,
ali@186	171	"Enable line end checking", NULL },
ali@186	172	{ "no-line-end", 'l', G_OPTION_FLAG_REVERSE,
ali@186	173	G_OPTION_ARG_NONE, pswit+LINE_END_SWITCH,
ali@186	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@186	177	{ "no-overview", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	178	G_OPTION_ARG_NONE, pswit+OVERVIEW_SWITCH,
ali@186	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@186	182	{ "no-stdout", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	183	G_OPTION_ARG_NONE, pswit+STDOUT_SWITCH,
ali@186	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@186	187	{ "no-header", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	188	G_OPTION_ARG_NONE, pswit+HEADER_SWITCH,
ali@186	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@186	192	{ "no-markup", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	193	G_OPTION_ARG_NONE, pswit+MARKUP_SWITCH,
ali@186	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@186	197	{ "no-usertypo", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	198	G_OPTION_ARG_NONE, pswit+USERTYPO_SWITCH,
ali@186	199	"Ignore file of user-defined typos", NULL },
ali@186	200	{ "verbose", 'v', 0, G_OPTION_ARG_NONE, pswit+VERBOSE_SWITCH,
ali@186	201	"Verbose - list everything", NULL },
ali@186	202	{ "no-verbose", 0, G_OPTION_FLAG_HIDDEN\|G_OPTION_FLAG_REVERSE,
ali@186	203	G_OPTION_ARG_NONE, pswit+VERBOSE_SWITCH,
ali@186	204	"Switch off verbose mode", NULL },
ali@192	205	{ "charset", 0, 0, G_OPTION_ARG_STRING, &opt_charset,
ali@192	206	"Set of characters valid for this ebook", "NAME" },
ali@186	207	{ NULL }
ali@186	208	};
ali@186	209
ali@186	210	/*
ali@186	211	* Options relating to configuration which make no sense from inside
ali@186	212	* a configuration file.
ali@186	213	*/
ali@186	214
ali@186	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@186	218	{ "dump-config", 0, 0, G_OPTION_ARG_NONE, pswit+DUMP_CONFIG_SWITCH,
ali@186	219	"Dump current config settings", NULL },
ali@186	220	{ NULL }
ali@186	221	};
ali@186	222
ali@186	223	static GOptionEntry compatibility_options[]={
ali@186	224	{ "toggle-typo", 't', 0, G_OPTION_ARG_NONE, &typo_compat,
ali@186	225	"Toggle checking for common typos", NULL },
ali@186	226	{ "toggle-relaxed", 'x', 0, G_OPTION_ARG_NONE, &paranoid_compat,
ali@186	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@186	271	GKeyFile *config;
ali@186	272
ali@186	273	void config_file_update(GKeyFile *kf)
ali@186	274	{
ali@186	275	int i;
ali@186	276	gboolean sw;
ali@186	277	for(i=0;options[i].long_name;i++)
ali@186	278	{
ali@186	279	if (g_str_has_prefix(options[i].long_name,"no-"))
ali@186	280	continue;
ali@186	281	if (options[i].arg==G_OPTION_ARG_NONE)
ali@186	282	{
ali@186	283	sw=(gboolean )options[i].arg_data;
ali@186	284	if (options[i].flags&G_OPTION_FLAG_REVERSE)
ali@186	285	sw=!sw;
ali@186	286	g_key_file_set_boolean(kf,"options",options[i].long_name,sw);
ali@186	287	}
ali@186	288	else
ali@186	289	g_assert_not_reached();
ali@186	290	}
ali@186	291	}
ali@186	292
ali@186	293	void config_file_add_comments(GKeyFile *kf)
ali@186	294	{
ali@186	295	int i;
ali@186	296	gchar *comment;
ali@186	297	g_key_file_set_comment(kf,NULL,NULL," Default configuration for bookloupe",
ali@186	298	NULL);
ali@186	299	for(i=0;options[i].long_name;i++)
ali@186	300	{
ali@186	301	if (g_str_has_prefix(options[i].long_name,"no-"))
ali@186	302	continue;
ali@186	303	comment=g_strconcat(" ",options[i].description,NULL);
ali@186	304	g_key_file_set_comment(kf,"options",options[i].long_name,comment,NULL);
ali@186	305	g_free(comment);
ali@186	306	}
ali@186	307	}
ali@186	308
ali@186	309	void dump_config(void)
ali@186	310	{
ali@186	311	gchar *s;
ali@186	312	if (config)
ali@186	313	config_file_update(config);
ali@186	314	else
ali@186	315	{
ali@186	316	config=g_key_file_new();
ali@186	317	config_file_update(config);
ali@186	318	config_file_add_comments(config);
ali@186	319	}
ali@186	320	s=g_key_file_to_data(config,NULL,NULL);
ali@186	321	if (s)
ali@186	322	g_print("%s",s);
ali@186	323	g_free(s);
ali@186	324	}
ali@186	325
ali@186	326	GKeyFile read_config_file(gchar *full_path)
ali@186	327	{
ali@186	328	int i;
ali@186	329	GError *err=NULL;
ali@186	330	gchar **search_dirs;
ali@186	331	gchar *path;
ali@186	332	const char *search_path;
ali@186	333	GKeyFile *kf;
ali@186	334	kf=g_key_file_new();
ali@186	335	search_path=g_getenv("BOOKLOUPE_CONFIG_PATH");
ali@186	336	if (search_path)
ali@186	337	{
ali@186	338	#ifdef __WIN32__
ali@186	339	search_dirs=g_strsplit(search_path,";",0);
ali@186	340	#else
ali@186	341	search_dirs=g_strsplit(search_path,":",0);
ali@186	342	#endif
ali@186	343	}
ali@186	344	else
ali@186	345	{
ali@186	346	search_dirs=g_new(gchar *,4);
ali@186	347	search_dirs[0]=g_get_current_dir();
ali@186	348	search_dirs[1]=g_strdup(running_from);
ali@186	349	search_dirs[2]=g_strdup(g_get_user_config_dir());
ali@186	350	search_dirs[3]=NULL;
ali@186	351	}
ali@186	352	for(i=0;search_dirs[i];i++)
ali@186	353	{
ali@186	354	path=g_build_filename(search_dirs[i],"bookloupe.ini",NULL);
ali@186	355	if (g_key_file_load_from_file(kf,path,
ali@186	356	G_KEY_FILE_KEEP_COMMENTS\|G_KEY_FILE_KEEP_TRANSLATIONS,&err))
ali@186	357	break;
ali@186	358	if (!g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@186	359	{
ali@186	360	g_printerr("Bookloupe: Error reading %s\n",path);
ali@186	361	g_printerr("%s\n",err->message);
ali@186	362	exit(1);
ali@186	363	}
ali@186	364	g_clear_error(&err);
ali@186	365	g_free(path);
ali@186	366	path=NULL;
ali@186	367	}
ali@186	368	if (!search_dirs[i])
ali@186	369	{
ali@186	370	g_key_file_free(kf);
ali@186	371	kf=NULL;
ali@186	372	}
ali@186	373	g_strfreev(search_dirs);
ali@186	374	if (full_path && kf)
ali@186	375	*full_path=path;
ali@186	376	else
ali@186	377	g_free(path);
ali@186	378	return kf;
ali@186	379	}
ali@186	380
ali@186	381	void parse_config_file(void)
ali@186	382	{
ali@186	383	int i,j;
ali@186	384	gchar *path;
ali@186	385	gchar **keys;
ali@186	386	gboolean sw;
ali@186	387	GError *err=NULL;
ali@186	388	config=read_config_file(&path);
ali@186	389	if (config)
ali@186	390	keys=g_key_file_get_keys(config,"options",NULL,NULL);
ali@186	391	else
ali@186	392	keys=NULL;
ali@186	393	if (keys)
ali@186	394	{
ali@186	395	for(i=0;keys[i];i++)
ali@186	396	{
ali@186	397	for(j=0;options[j].long_name;j++)
ali@186	398	{
ali@186	399	if (g_str_has_prefix(options[j].long_name,"no-"))
ali@186	400	continue;
ali@186	401	else if (!strcmp(keys[i],options[j].long_name))
ali@186	402	{
ali@186	403	if (options[j].arg==G_OPTION_ARG_NONE)
ali@186	404	{
ali@186	405	sw=g_key_file_get_boolean(config,"options",keys[i],
ali@186	406	&err);
ali@186	407	if (err)
ali@186	408	{
ali@186	409	g_printerr("Bookloupe: %s: options.%s: %s\n",
ali@186	410	path,keys[i],err->message);
ali@186	411	g_clear_error(&err);
ali@186	412	}
ali@186	413	if (options[j].flags&G_OPTION_FLAG_REVERSE)
ali@186	414	sw=!sw;
ali@186	415	(gboolean )options[j].arg_data=sw;
ali@186	416	break;
ali@186	417	}
ali@186	418	else
ali@186	419	g_assert_not_reached();
ali@186	420	}
ali@186	421	}
ali@186	422	if (!options[j].long_name)
ali@186	423	g_printerr("Bookloupe: %s: Unknown option \"%s\" ignored\n",
ali@186	424	path,keys[i]);
ali@186	425	}
ali@186	426	g_strfreev(keys);
ali@186	427	}
ali@186	428	if (config)
ali@186	429	g_free(path);
ali@186	430	}
ali@186	431
ali@185	432	gboolean set_charset(const char name,GError *err)
ali@185	433	{
ali@185	434	/* The various UNICODE encodings all share the same character set. */
ali@185	435	const char *unicode_aliases[]={ "UCS-2", "UCS-2BE", "UCS-2LE", "UCS-4",
ali@185	436	"UCS-4BE", "UCS-4LE", "UCS2", "UCS4", "UNICODE", "UNICODEBIG",
ali@185	437	"UNICODELITTLE", "UTF-7", "UTF-8", "UTF-16", "UTF-16BE", "UTF-16LE",
ali@185	438	"UTF-32", "UTF-32BE", "UTF-32LE", "UTF7", "UTF8", "UTF16", "UTF16BE",
ali@185	439	"UTF16LE", "UTF32", "UTF32BE", "UTF32LE" };
ali@185	440	int i;
ali@185	441	if (charset)
ali@185	442	g_free(charset);
ali@185	443	if (charset_validator!=(GIConv)-1)
ali@185	444	g_iconv_close(charset_validator);
ali@185	445	if (!name \|\| !g_strcasecmp(name,"auto"))
ali@185	446	{
ali@185	447	charset=NULL;
ali@185	448	charset_validator=(GIConv)-1;
ali@185	449	return TRUE;
ali@185	450	}
ali@185	451	else
ali@185	452	charset=g_strdup(name);
ali@185	453	for(i=0;i<G_N_ELEMENTS(unicode_aliases);i++)
ali@185	454	if (!g_strcasecmp(charset,unicode_aliases[i]))
ali@185	455	{
ali@185	456	g_free(charset);
ali@185	457	charset=g_strdup("UTF-8");
ali@185	458	break;
ali@185	459	}
ali@185	460	if (!strcmp(charset,"UTF-8"))
ali@185	461	charset_validator=(GIConv)-1;
ali@185	462	else
ali@185	463	{
ali@185	464	charset_validator=g_iconv_open(charset,"UTF-8");
ali@185	465	if (charset_validator==(GIConv)-1)
ali@185	466	{
ali@185	467	g_set_error(err,G_CONVERT_ERROR,G_CONVERT_ERROR_NO_CONVERSION,
ali@185	468	"Unknown character set \"%s\"",charset);
ali@185	469	return FALSE;
ali@185	470	}
ali@185	471	}
ali@185	472	return TRUE;
ali@185	473	}
ali@185	474
ali@69	475	void parse_options(int argc,char **argv)
ali@0	476	{
ali@69	477	GError *err=NULL;
ali@69	478	GOptionContext *context;
ali@186	479	GOptionGroup *compatibility;
ali@69	480	context=g_option_context_new(
ali@186	481	"file - look for errors in Project Gutenberg(TM) etexts");
ali@69	482	g_option_context_add_main_entries(context,options,NULL);
ali@186	483	g_option_context_add_main_entries(context,config_options,NULL);
ali@186	484	compatibility=g_option_group_new("compatibility",
ali@186	485	"Options for Compatibility with Gutcheck:",
ali@186	486	"Show compatibility options",NULL,NULL);
ali@186	487	g_option_group_add_entries(compatibility,compatibility_options);
ali@186	488	g_option_context_add_group(context,compatibility);
ali@186	489	g_option_context_set_description(context,
ali@186	490	"For simplicity, only the switch options which reverse the\n"
ali@186	491	"default configuration are listed. In most cases, both vanilla\n"
ali@186	492	"and \"no-\" prefixed versions are available for use.");
ali@69	493	if (!g_option_context_parse(context,argc,argv,&err))
ali@69	494	{
ali@69	495	g_printerr("Bookloupe: %s\n",err->message);
ali@69	496	g_printerr("Use \"%s --help\" for help\n",(*argv)[0]);
ali@69	497	exit(1);
ali@69	498	}
ali@186	499	if (typo_compat)
ali@69	500	pswit[TYPO_SWITCH]=!pswit[TYPO_SWITCH];
ali@186	501	if (paranoid_compat)
ali@186	502	{
ali@186	503	pswit[PARANOID_SWITCH]=!pswit[PARANOID_SWITCH];
ali@186	504	pswit[TYPO_SWITCH]=!pswit[TYPO_SWITCH];
ali@186	505	}
ali@40	506	/*
ali@40	507	* Web uploads - for the moment, this is really just a placeholder
ali@40	508	* until we decide what processing we really want to do on web uploads
ali@40	509	*/
ali@40	510	if (pswit[WEB_SWITCH])
ali@40	511	{
ali@40	512	/* specific override for web uploads */
ali@69	513	pswit[ECHO_SWITCH]=TRUE;
ali@69	514	pswit[SQUOTE_SWITCH]=FALSE;
ali@69	515	pswit[TYPO_SWITCH]=TRUE;
ali@69	516	pswit[QPARA_SWITCH]=FALSE;
ali@69	517	pswit[PARANOID_SWITCH]=TRUE;
ali@69	518	pswit[LINE_END_SWITCH]=FALSE;
ali@69	519	pswit[OVERVIEW_SWITCH]=FALSE;
ali@69	520	pswit[STDOUT_SWITCH]=FALSE;
ali@69	521	pswit[HEADER_SWITCH]=TRUE;
ali@69	522	pswit[VERBOSE_SWITCH]=FALSE;
ali@69	523	pswit[MARKUP_SWITCH]=FALSE;
ali@69	524	pswit[USERTYPO_SWITCH]=FALSE;
ali@69	525	pswit[DP_SWITCH]=FALSE;
ali@40	526	}
ali@185	527	if (opt_charset && !set_charset(opt_charset,&err))
ali@185	528	{
ali@185	529	g_printerr("%s\n",err->message);
ali@185	530	exit(1);
ali@185	531	}
ali@186	532	if (pswit[DUMP_CONFIG_SWITCH])
ali@186	533	{
ali@186	534	dump_config();
ali@186	535	exit(0);
ali@186	536	}
ali@185	537	g_free(opt_charset);
ali@185	538	opt_charset=NULL;
ali@186	539	if (pswit[OVERVIEW_SWITCH])
ali@186	540	/* just print summary; don't echo */
ali@186	541	pswit[ECHO_SWITCH]=FALSE;
ali@69	542	if (*argc<2)
ali@40	543	{
ali@69	544	proghelp(context);
ali@69	545	exit(1);
ali@40	546	}
ali@69	547	g_option_context_free(context);
ali@69	548	}
ali@69	549
ali@69	550	/*
ali@69	551	* read_user_scannos:
ali@69	552	*
ali@69	553	* Read in the user-defined stealth scanno list.
ali@69	554	*/
ali@69	555	void read_user_scannos(void)
ali@69	556	{
ali@69	557	GError *err=NULL;
ali@69	558	gchar *usertypo_file;
ali@69	559	gboolean okay;
ali@69	560	int i;
ali@70	561	gsize len,nb;
ali@70	562	gchar contents,utf8,**lines;
ali@69	563	usertypo_file=g_strdup("bookloupe.typ");
ali@69	564	okay=file_get_contents_text(usertypo_file,&contents,&len,&err);
ali@69	565	if (g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@69	566	{
ali@69	567	g_clear_error(&err);
ali@69	568	g_free(usertypo_file);
ali@69	569	usertypo_file=g_build_filename(running_from,"bookloupe.typ",NULL);
ali@69	570	okay=file_get_contents_text(usertypo_file,&contents,&len,&err);
ali@69	571	}
ali@69	572	if (g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@69	573	{
ali@69	574	g_clear_error(&err);
ali@69	575	g_free(usertypo_file);
ali@69	576	usertypo_file=g_strdup("gutcheck.typ");
ali@69	577	okay=file_get_contents_text(usertypo_file,&contents,&len,&err);
ali@69	578	}
ali@69	579	if (g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@69	580	{
ali@69	581	g_clear_error(&err);
ali@69	582	g_free(usertypo_file);
ali@69	583	usertypo_file=g_build_filename(running_from,"gutcheck.typ",NULL);
ali@69	584	okay=file_get_contents_text(usertypo_file,&contents,&len,&err);
ali@69	585	}
ali@69	586	if (g_error_matches(err,G_FILE_ERROR,G_FILE_ERROR_NOENT))
ali@69	587	{
ali@69	588	g_free(usertypo_file);
ali@70	589	g_print(" --> I couldn't find bookloupe.typ "
ali@69	590	"-- proceeding without user typos.\n");
ali@69	591	return;
ali@69	592	}
ali@69	593	else if (!okay)
ali@69	594	{
ali@69	595	fprintf(stderr,"%s: %s\n",usertypo_file,err->message);
ali@69	596	g_free(usertypo_file);
ali@69	597	g_clear_error(&err);
ali@69	598	exit(1);
ali@69	599	}
ali@72	600	if (g_utf8_validate(contents,len,NULL))
ali@185	601	{
ali@72	602	utf8=g_utf8_normalize(contents,len,G_NORMALIZE_DEFAULT_COMPOSE);
ali@185	603	if (!charset)
ali@185	604	(void)set_charset("UNICODE",NULL);
ali@185	605	}
ali@72	606	else
ali@72	607	utf8=g_convert(contents,len,"UTF-8","WINDOWS-1252",NULL,&nb,NULL);
ali@70	608	g_free(contents);
ali@70	609	lines=g_strsplit_set(utf8,"\r\n",0);
ali@70	610	g_free(utf8);
ali@69	611	usertypo=g_tree_new_full((GCompareDataFunc)strcmp,NULL,g_free,NULL);
ali@69	612	for (i=0;lines[i];i++)
ali@69	613	if ((unsigned char )lines[i]>'!')
ali@69	614	g_tree_insert(usertypo,lines[i],GINT_TO_POINTER(1));
ali@69	615	else
ali@69	616	g_free(lines[i]);
ali@69	617	g_free(lines);
ali@69	618	}
ali@69	619
ali@69	620	/*
ali@69	621	* read_etext:
ali@69	622	*
ali@69	623	* Read an etext returning a newly allocated string containing the file
ali@69	624	* contents or NULL on error.
ali@69	625	*/
ali@69	626	gchar read_etext(const char filename,GError **err)
ali@69	627	{
ali@76	628	GError *tmp_err=NULL;
ali@70	629	gchar contents,utf8;
ali@76	630	gsize len,bytes_read,bytes_written;
ali@76	631	int i,line,col;
ali@69	632	if (!g_file_get_contents(filename,&contents,&len,err))
ali@69	633	return NULL;
ali@72	634	if (g_utf8_validate(contents,len,NULL))
ali@72	635	{
ali@72	636	utf8=g_utf8_normalize(contents,len,G_NORMALIZE_DEFAULT_COMPOSE);
ali@72	637	g_set_print_handler(print_as_utf_8);
ali@73	638	#ifdef __WIN32__
ali@73	639	SetConsoleOutputCP(CP_UTF8);
ali@73	640	#endif
ali@72	641	}
ali@72	642	else
ali@72	643	{
ali@76	644	utf8=g_convert(contents,len,"UTF-8","WINDOWS-1252",&bytes_read,
ali@76	645	&bytes_written,&tmp_err);
ali@76	646	if (g_error_matches(tmp_err,G_CONVERT_ERROR,
ali@76	647	G_CONVERT_ERROR_ILLEGAL_SEQUENCE))
ali@76	648	{
ali@76	649	line=col=1;
ali@76	650	for(i=0;i<bytes_read;i++)
ali@76	651	if (contents[i]=='\n')
ali@76	652	{
ali@76	653	line++;
ali@76	654	col=1;
ali@76	655	}
ali@76	656	else if (contents[i]!='\r')
ali@76	657	col++;
ali@76	658	g_set_error(err,G_CONVERT_ERROR,G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
ali@76	659	"Input conversion failed. Byte %d at line %d, column %d is not a "
ali@76	660	"valid Windows-1252 character",
ali@76	661	((unsigned char *)contents)[bytes_read],line,col);
ali@76	662	}
ali@76	663	else if (tmp_err)
ali@76	664	g_propagate_error(err,tmp_err);
ali@72	665	g_set_print_handler(print_as_windows_1252);
ali@73	666	#ifdef __WIN32__
ali@73	667	SetConsoleOutputCP(1252);
ali@73	668	#endif
ali@72	669	}
ali@70	670	g_free(contents);
ali@70	671	return utf8;
ali@69	672	}
ali@69	673
ali@73	674	void cleanup_on_exit(void)
ali@73	675	{
ali@73	676	#ifdef __WIN32__
ali@73	677	SetConsoleOutputCP(saved_cp);
ali@73	678	#endif
ali@73	679	}
ali@73	680
ali@69	681	int main(int argc,char **argv)
ali@69	682	{
ali@73	683	#ifdef __WIN32__
ali@73	684	atexit(cleanup_on_exit);
ali@73	685	saved_cp=GetConsoleOutputCP();
ali@73	686	#endif
ali@69	687	running_from=g_path_get_dirname(argv[0]);
ali@186	688	/* Paranoid checking is turned OFF, not on, by its switch */
ali@186	689	pswit[PARANOID_SWITCH]=TRUE;
ali@186	690	/* if running in paranoid mode, typo checks default to enabled */
ali@186	691	pswit[TYPO_SWITCH]=TRUE;
ali@186	692	/* Line-end checking is turned OFF, not on, by its switch */
ali@186	693	pswit[LINE_END_SWITCH]=TRUE;
ali@186	694	/* Echoing is turned OFF, not on, by its switch */
ali@186	695	pswit[ECHO_SWITCH]=TRUE;
ali@186	696	parse_config_file();
ali@69	697	parse_options(&argc,&argv);
ali@40	698	if (pswit[USERTYPO_SWITCH])
ali@69	699	read_user_scannos();
ali@40	700	fprintf(stderr,"bookloupe: Check and report on an e-text\n");
ali@69	701	procfile(argv[1]);
ali@40	702	if (pswit[OVERVIEW_SWITCH])
ali@40	703	{
ali@70	704	g_print(" Checked %ld lines of %ld (head+foot = %ld)\n\n",
ali@40	705	checked_linecnt,linecnt,linecnt-checked_linecnt);
ali@70	706	g_print(" --------------- Queries found --------------\n");
ali@68	707	if (cnt_long)
ali@70	708	g_print(" Long lines: %14ld\n",cnt_long);
ali@68	709	if (cnt_short)
ali@70	710	g_print(" Short lines: %14ld\n",cnt_short);
ali@68	711	if (cnt_lineend)
ali@70	712	g_print(" Line-end problems: %14ld\n",cnt_lineend);
ali@68	713	if (cnt_word)
ali@70	714	g_print(" Common typos: %14ld\n",cnt_word);
ali@142	715	if (cnt_quote)
ali@142	716	g_print(" Unmatched quotes: %14ld\n",cnt_quote);
ali@68	717	if (cnt_brack)
ali@70	718	g_print(" Unmatched brackets: %14ld\n",cnt_brack);
ali@68	719	if (cnt_bin)
ali@70	720	g_print(" Non-ASCII characters: %14ld\n",cnt_bin);
ali@68	721	if (cnt_odd)
ali@70	722	g_print(" Proofing characters: %14ld\n",cnt_odd);
ali@68	723	if (cnt_punct)
ali@70	724	g_print(" Punctuation & spacing queries: %14ld\n",cnt_punct);
ali@68	725	if (cnt_dash)
ali@70	726	g_print(" Non-standard dashes: %14ld\n",cnt_dash);
ali@68	727	if (cnt_html)
ali@70	728	g_print(" Possible HTML tags: %14ld\n",cnt_html);
ali@70	729	g_print("\n");
ali@70	730	g_print(" TOTAL QUERIES %14ld\n",
ali@142	731	cnt_quote+cnt_brack+cnt_bin+cnt_odd+cnt_long+cnt_short+cnt_punct+
ali@142	732	cnt_dash+cnt_word+cnt_html+cnt_lineend);
ali@40	733	}
ali@69	734	g_free(running_from);
ali@69	735	if (usertypo)
ali@69	736	g_tree_unref(usertypo);
ali@185	737	set_charset(NULL,NULL);
ali@186	738	if (config)
ali@186	739	g_key_file_free(config);
ali@40	740	return 0;
ali@0	741	}
ali@0	742
ali@147	743	void count_dashes(const char line,const char dash,
ali@147	744	struct dash_results *results)
ali@147	745	{
ali@147	746	int i;
ali@147	747	gchar **tokens;
ali@147	748	gunichar pc,nc;
ali@147	749	gboolean spaced=FALSE,unspaced=FALSE,spaced2=FALSE;
ali@147	750	if (!*line)
ali@147	751	return;
ali@147	752	tokens=g_strsplit(line,dash,0);
ali@147	753	if (tokens[1])
ali@147	754	results->base++;
ali@147	755	for(i=1;tokens[i];i++)
ali@147	756	{
ali@147	757	pc=g_utf8_get_char(g_utf8_prev_char(tokens[i-1]+strlen(tokens[i-1])));
ali@147	758	nc=g_utf8_get_char(tokens[i]);
ali@147	759	if (g_unichar_isspace(pc) \|\| g_unichar_isspace(nc))
ali@147	760	spaced=TRUE;
ali@147	761	if (g_unichar_isspace(pc) && g_unichar_isspace(nc))
ali@147	762	spaced2=TRUE;
ali@147	763	else if (!g_unichar_isspace(pc) && !g_unichar_isspace(nc))
ali@147	764	unspaced=TRUE;
ali@147	765	}
ali@147	766	if (spaced)
ali@147	767	results->space++;
ali@147	768	if (spaced2)
ali@147	769	/* count of lines with em-dashes with spaces both sides */
ali@147	770	results->non_PG_space++;
ali@147	771	if (unspaced)
ali@147	772	/* count of lines with PG-type em-dashes with no spaces */
ali@147	773	results->PG_space++;
ali@147	774	g_strfreev(tokens);
ali@147	775	}
ali@147	776
ali@40	777	/*
ali@41	778	* first_pass:
ali@40	779	*
ali@41	780	* Run a first pass - verify that it's a valid PG
ali@41	781	* file, decide whether to report some things that
ali@41	782	* occur many times in the text like long or short
ali@41	783	* lines, non-standard dashes, etc.
ali@40	784	*/
ali@69	785	struct first_pass_results first_pass(const char etext)
ali@0	786	{
ali@70	787	gunichar laststart=CHAR_SPACE;
ali@54	788	const char *s;
ali@69	789	gchar *lc_line;
ali@70	790	int i,j,lbytes,llen;
ali@69	791	gchar **lines;
ali@41	792	unsigned int lastlen=0,lastblen=0;
ali@41	793	long spline=0,nspline=0;
ali@41	794	static struct first_pass_results results={0};
ali@147	795	struct dash_results tmp_dash_results;
ali@69	796	gchar *inword;
ali@142	797	QuoteClass qc;
ali@69	798	lines=g_strsplit(etext,"\n",0);
ali@69	799	for (j=0;lines[j];j++)
ali@40	800	{
ali@70	801	lbytes=strlen(lines[j]);
ali@82	802	while (lbytes>0 && lines[j][lbytes-1]=='\r')
ali@70	803	lines[j][--lbytes]='\0';
ali@70	804	llen=g_utf8_strlen(lines[j],lbytes);
ali@68	805	linecnt++;
ali@69	806	if (strstr(lines[j],"*END") && strstr(lines[j],"SMALL PRINT") &&
ali@69	807	(strstr(lines[j],"PUBLIC DOMAIN") \|\| strstr(lines[j],"COPYRIGHT")))
ali@40	808	{
ali@68	809	if (spline)
ali@70	810	g_print(" --> Duplicate header?\n");
ali@68	811	spline=linecnt+1; /* first line of non-header text, that is */
ali@40	812	}
ali@69	813	if (!strncmp(lines[j],"*** START",9) &&
ali@69	814	strstr(lines[j],"PROJECT GUTENBERG"))
ali@40	815	{
ali@68	816	if (nspline)
ali@70	817	g_print(" --> Duplicate header?\n");
ali@68	818	nspline=linecnt+1; /* first line of non-header text, that is */
ali@40	819	}
ali@68	820	if (spline \|\| nspline)
ali@40	821	{
ali@70	822	lc_line=g_utf8_strdown(lines[j],lbytes);
ali@69	823	if (strstr(lc_line,"end") && strstr(lc_line,"project gutenberg"))
ali@40	824	{
ali@69	825	if (strstr(lc_line,"end")<strstr(lc_line,"project gutenberg"))
ali@40	826	{
ali@68	827	if (results.footerline)
ali@40	828	{
ali@40	829	/* it's an old-form header - we can detect duplicates */
ali@68	830	if (!nspline)
ali@70	831	g_print(" --> Duplicate footer?\n");
ali@40	832	}
ali@68	833	else
ali@68	834	results.footerline=linecnt;
ali@40	835	}
ali@40	836	}
ali@69	837	g_free(lc_line);
ali@40	838	}
ali@68	839	if (spline)
ali@41	840	results.firstline=spline;
ali@68	841	if (nspline)
ali@41	842	results.firstline=nspline; /* override with new */
ali@68	843	if (results.footerline)
ali@40	844	continue; /* don't count the boilerplate in the footer */
ali@68	845	results.totlen+=llen;
ali@70	846	for (s=lines[j];*s;s=g_utf8_next_char(s))
ali@40	847	{
ali@70	848	if (g_utf8_get_char(s)>127)
ali@41	849	results.binlen++;
ali@70	850	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@41	851	results.alphalen++;
ali@142	852	if (s>lines[j])
ali@142	853	{
ali@142	854	if (CHAR_IS_DQUOTE(g_utf8_get_char(s)))
ali@142	855	qc=QUOTE_CLASS(g_utf8_get_char(s));
ali@142	856	else
ali@142	857	qc=INVALID_QUOTE;
ali@142	858	if ((qc==CLOSING_QUOTE \|\| qc==NEUTRAL_QUOTE) &&
ali@147	859	g_unichar_isalpha(g_utf8_get_char(g_utf8_prev_char(s))))
ali@142	860	results.endquote_count++;
ali@142	861	}
ali@40	862	}
ali@69	863	if (llen>2 && lastlen>2 && lastlen<SHORTEST_PG_LINE && lastblen>2 &&
ali@69	864	lastblen>SHORTEST_PG_LINE && laststart!=CHAR_SPACE)
ali@41	865	results.shortline++;
ali@70	866	if (lbytes>0 &&
ali@70	867	g_utf8_get_char(g_utf8_prev_char(lines[j]+lbytes))<=CHAR_SPACE)
ali@40	868	cnt_spacend++;
ali@69	869	if (strstr(lines[j],".,"))
ali@41	870	results.dotcomma++;
ali@68	871	/* only count ast lines for ignoring purposes where there is */
ali@68	872	/* locase text on the line */
ali@69	873	if (strchr(lines[j],'*'))
ali@40	874	{
ali@70	875	for (s=lines[j];*s;s=g_utf8_next_char(s))
ali@70	876	if (g_unichar_islower(g_utf8_get_char(s)))
ali@68	877	break;
ali@70	878	if (*s)
ali@41	879	results.astline++;
ali@40	880	}
ali@69	881	if (strchr(lines[j],'/'))
ali@68	882	results.fslashline++;
ali@82	883	if (lbytes>0)
ali@82	884	{
ali@82	885	for (s=g_utf8_prev_char(lines[j]+lbytes);
ali@82	886	s>lines[j] && g_utf8_get_char(s)<=CHAR_SPACE;
ali@82	887	s=g_utf8_prev_char(s))
ali@82	888	;
ali@82	889	if (s>g_utf8_next_char(lines[j]) && g_utf8_get_char(s)=='-' &&
ali@82	890	g_utf8_get_char(g_utf8_prev_char(s))!='-')
ali@82	891	results.hyphens++;
ali@82	892	}
ali@68	893	if (llen>LONGEST_PG_LINE)
ali@41	894	results.longline++;
ali@68	895	if (llen>WAY_TOO_LONG)
ali@41	896	results.verylongline++;
ali@69	897	if (strchr(lines[j],'<') && strchr(lines[j],'>'))
ali@40	898	{
ali@69	899	i=(int)(strchr(lines[j],'>')-strchr(lines[j],'<')+1);
ali@68	900	if (i>0)
ali@68	901	results.htmcount++;
ali@69	902	if (strstr(lines[j],"<i>"))
ali@41	903	results.htmcount+=4; /* bonus marks! */
ali@40	904	}
ali@68	905	/* Check for spaced em-dashes */
ali@147	906	memset(&tmp_dash_results,0,sizeof(tmp_dash_results));
ali@147	907	count_dashes(lines[j],"--",&tmp_dash_results);
ali@147	908	count_dashes(lines[j],"—",&tmp_dash_results);
ali@147	909	if (tmp_dash_results.base)
ali@147	910	results.emdash.base++;
ali@147	911	if (tmp_dash_results.non_PG_space)
ali@147	912	results.emdash.non_PG_space++;
ali@147	913	if (tmp_dash_results.PG_space)
ali@147	914	results.emdash.PG_space++;
ali@69	915	for (s=lines[j];*s;)
ali@40	916	{
ali@69	917	inword=getaword(&s);
ali@68	918	if (!strcmp(inword,"hij") \|\| !strcmp(inword,"niet"))
ali@68	919	results.Dutchcount++;
ali@68	920	if (!strcmp(inword,"dans") \|\| !strcmp(inword,"avec"))
ali@68	921	results.Frenchcount++;
ali@68	922	if (!strcmp(inword,"0") \|\| !strcmp(inword,"1"))
ali@68	923	results.standalone_digit++;
ali@69	924	g_free(inword);
ali@40	925	}
ali@68	926	/* Check for spaced dashes */
ali@69	927	if (strstr(lines[j]," -") && *(strstr(lines[j]," -")+2)!='-')
ali@41	928	results.spacedash++;
ali@68	929	lastblen=lastlen;
ali@69	930	lastlen=llen;
ali@69	931	laststart=lines[j][0];
ali@40	932	}
ali@69	933	g_strfreev(lines);
ali@41	934	return &results;
ali@41	935	}
ali@41	936
ali@42	937	/*
ali@42	938	* report_first_pass:
ali@42	939	*
ali@42	940	* Make some snap decisions based on the first pass results.
ali@42	941	*/
ali@42	942	struct warnings report_first_pass(struct first_pass_results results)
ali@42	943	{
ali@42	944	static struct warnings warnings={0};
ali@42	945	if (cnt_spacend>0)
ali@70	946	g_print(" --> %ld lines in this file have white space at end\n",
ali@42	947	cnt_spacend);
ali@42	948	warnings.dotcomma=1;
ali@42	949	if (results->dotcomma>5)
ali@42	950	{
ali@68	951	warnings.dotcomma=0;
ali@70	952	g_print(" --> %ld lines in this file contain '.,'. "
ali@42	953	"Not reporting them.\n",results->dotcomma);
ali@42	954	}
ali@42	955	/*
ali@42	956	* If more than 50 lines, or one-tenth, are short,
ali@42	957	* don't bother reporting them.
ali@42	958	*/
ali@42	959	warnings.shortline=1;
ali@42	960	if (results->shortline>50 \|\| results->shortline*10>linecnt)
ali@42	961	{
ali@68	962	warnings.shortline=0;
ali@70	963	g_print(" --> %ld lines in this file are short. "
ali@42	964	"Not reporting short lines.\n",results->shortline);
ali@42	965	}
ali@42	966	/*
ali@42	967	* If more than 50 lines, or one-tenth, are long,
ali@42	968	* don't bother reporting them.
ali@42	969	*/
ali@42	970	warnings.longline=1;
ali@42	971	if (results->longline>50 \|\| results->longline*10>linecnt)
ali@42	972	{
ali@68	973	warnings.longline=0;
ali@70	974	g_print(" --> %ld lines in this file are long. "
ali@42	975	"Not reporting long lines.\n",results->longline);
ali@42	976	}
ali@42	977	/* If more than 10 lines contain asterisks, don't bother reporting them. */
ali@42	978	warnings.ast=1;
ali@42	979	if (results->astline>10)
ali@42	980	{
ali@68	981	warnings.ast=0;
ali@70	982	g_print(" --> %ld lines in this file contain asterisks. "
ali@42	983	"Not reporting them.\n",results->astline);
ali@42	984	}
ali@42	985	/*
ali@42	986	* If more than 10 lines contain forward slashes,
ali@42	987	* don't bother reporting them.
ali@42	988	*/
ali@42	989	warnings.fslash=1;
ali@42	990	if (results->fslashline>10)
ali@42	991	{
ali@68	992	warnings.fslash=0;
ali@70	993	g_print(" --> %ld lines in this file contain forward slashes. "
ali@42	994	"Not reporting them.\n",results->fslashline);
ali@42	995	}
ali@42	996	/*
ali@42	997	* If more than 20 lines contain unpunctuated endquotes,
ali@42	998	* don't bother reporting them.
ali@42	999	*/
ali@42	1000	warnings.endquote=1;
ali@42	1001	if (results->endquote_count>20)
ali@42	1002	{
ali@68	1003	warnings.endquote=0;
ali@70	1004	g_print(" --> %ld lines in this file contain unpunctuated endquotes. "
ali@42	1005	"Not reporting them.\n",results->endquote_count);
ali@42	1006	}
ali@42	1007	/*
ali@42	1008	* If more than 15 lines contain standalone digits,
ali@42	1009	* don't bother reporting them.
ali@42	1010	*/
ali@42	1011	warnings.digit=1;
ali@42	1012	if (results->standalone_digit>10)
ali@42	1013	{
ali@68	1014	warnings.digit=0;
ali@70	1015	g_print(" --> %ld lines in this file contain standalone 0s and 1s. "
ali@42	1016	"Not reporting them.\n",results->standalone_digit);
ali@42	1017	}
ali@42	1018	/*
ali@42	1019	* If more than 20 lines contain hyphens at end,
ali@42	1020	* don't bother reporting them.
ali@42	1021	*/
ali@42	1022	warnings.hyphen=1;
ali@42	1023	if (results->hyphens>20)
ali@42	1024	{
ali@68	1025	warnings.hyphen=0;
ali@70	1026	g_print(" --> %ld lines in this file have hyphens at end. "
ali@42	1027	"Not reporting them.\n",results->hyphens);
ali@42	1028	}
ali@42	1029	if (results->htmcount>20 && !pswit[MARKUP_SWITCH])
ali@42	1030	{
ali@70	1031	g_print(" --> Looks like this is HTML. Switching HTML mode ON.\n");
ali@68	1032	pswit[MARKUP_SWITCH]=1;
ali@42	1033	}
ali@42	1034	if (results->verylongline>0)
ali@70	1035	g_print(" --> %ld lines in this file are VERY long!\n",
ali@42	1036	results->verylongline);
ali@42	1037	/*
ali@42	1038	* If there are more non-PG spaced dashes than PG em-dashes,
ali@42	1039	* assume it's deliberate.
ali@42	1040	* Current PG guidelines say don't use them, but older texts do,
ali@42	1041	* and some people insist on them whatever the guidelines say.
ali@42	1042	*/
ali@42	1043	warnings.dash=1;
ali@147	1044	if (results->spacedash+results->emdash.non_PG_space>
ali@147	1045	results->emdash.PG_space)
ali@42	1046	{
ali@68	1047	warnings.dash=0;
ali@70	1048	g_print(" --> There are %ld spaced dashes and em-dashes. "
ali@42	1049	"Not reporting them.\n",
ali@147	1050	results->spacedash+results->emdash.non_PG_space);
ali@42	1051	}
ali@185	1052	if (charset)
ali@185	1053	warnings.bin=0;
ali@185	1054	else
ali@42	1055	{
ali@185	1056	/* Charset ISO_8859-1/ASCII checks for compatibility with gutcheck */
ali@185	1057	warnings.bin=1;
ali@185	1058	/* If more than a quarter of characters are hi-bit, bug out. */
ali@185	1059	if (results->binlen*4>results->totlen)
ali@185	1060	{
ali@185	1061	g_print(" --> This file does not appear to be ASCII. "
ali@185	1062	"Terminating. Best of luck with it!\n");
ali@185	1063	exit(1);
ali@185	1064	}
ali@185	1065	if (results->alphalen*4<results->totlen)
ali@185	1066	{
ali@185	1067	g_print(" --> This file does not appear to be text. "
ali@185	1068	"Terminating. Best of luck with it!\n");
ali@185	1069	exit(1);
ali@185	1070	}
ali@185	1071	if (results->binlen*100>results->totlen \|\| results->binlen>100)
ali@185	1072	{
ali@185	1073	g_print(" --> There are a lot of foreign letters here. "
ali@185	1074	"Not reporting them.\n");
ali@185	1075	if (!pswit[VERBOSE_SWITCH])
ali@185	1076	warnings.bin=0;
ali@185	1077	}
ali@42	1078	}
ali@69	1079	warnings.isDutch=FALSE;
ali@42	1080	if (results->Dutchcount>50)
ali@42	1081	{
ali@69	1082	warnings.isDutch=TRUE;
ali@70	1083	g_print(" --> This looks like Dutch - "
ali@42	1084	"switching off dashes and warnings for 's Middags case.\n");
ali@42	1085	}
ali@69	1086	warnings.isFrench=FALSE;
ali@42	1087	if (results->Frenchcount>50)
ali@42	1088	{
ali@69	1089	warnings.isFrench=TRUE;
ali@70	1090	g_print(" --> This looks like French - "
ali@42	1091	"switching off some doublepunct.\n");
ali@42	1092	}
ali@42	1093	if (results->firstline && results->footerline)
ali@70	1094	g_print(" The PG header and footer appear to be already on.\n");
ali@42	1095	else
ali@42	1096	{
ali@68	1097	if (results->firstline)
ali@70	1098	g_print(" The PG header is on - no footer.\n");
ali@68	1099	if (results->footerline)
ali@70	1100	g_print(" The PG footer is on - no header.\n");
ali@42	1101	}
ali@70	1102	g_print("\n");
ali@42	1103	if (pswit[VERBOSE_SWITCH])
ali@42	1104	{
ali@68	1105	warnings.shortline=1;
ali@68	1106	warnings.dotcomma=1;
ali@68	1107	warnings.longline=1;
ali@68	1108	warnings.dash=1;
ali@68	1109	warnings.digit=1;
ali@68	1110	warnings.ast=1;
ali@68	1111	warnings.fslash=1;
ali@68	1112	warnings.hyphen=1;
ali@68	1113	warnings.endquote=1;
ali@70	1114	g_print(" * Verbose output is ON -- you asked for it! *\n");
ali@42	1115	}
ali@42	1116	if (warnings.isDutch)
ali@68	1117	warnings.dash=0;
ali@42	1118	if (results->footerline>0 && results->firstline>0 &&
ali@42	1119	results->footerline>results->firstline &&
ali@42	1120	results->footerline-results->firstline<100)
ali@42	1121	{
ali@70	1122	g_print(" --> I don't really know where this text starts. \n");
ali@70	1123	g_print(" There are no reference points.\n");
ali@70	1124	g_print(" I'm going to have to report the header and footer "
ali@42	1125	"as well.\n");
ali@68	1126	results->firstline=0;
ali@42	1127	}
ali@42	1128	return &warnings;
ali@42	1129	}
ali@42	1130
ali@43	1131	/*
ali@43	1132	* analyse_quotes:
ali@43	1133	*
ali@43	1134	* Look along the line, accumulate the count of quotes, and see
ali@43	1135	* if this is an empty line - i.e. a line with nothing on it
ali@43	1136	* but spaces.
ali@43	1137	* If line has just spaces, period, * and/or - on it, don't
ali@43	1138	* count it, since empty lines with asterisks or dashes to
ali@43	1139	* separate sections are common.
ali@43	1140	*
ali@69	1141	* Returns: TRUE if the line is empty.
ali@43	1142	*/
ali@164	1143	gboolean analyse_quotes(const char aline,struct counters counters)
ali@43	1144	{
ali@68	1145	int guessquote=0;
ali@69	1146	/* assume the line is empty until proven otherwise */
ali@69	1147	gboolean isemptyline=TRUE;
ali@70	1148	const char s=aline,sprev,*snext;
ali@70	1149	gunichar c;
ali@70	1150	sprev=NULL;
ali@142	1151	GError *tmp_err=NULL;
ali@43	1152	while (*s)
ali@43	1153	{
ali@70	1154	snext=g_utf8_next_char(s);
ali@70	1155	c=g_utf8_get_char(s);
ali@142	1156	if (CHAR_IS_DQUOTE(c))
ali@142	1157	(void)count_quote(counters,c,QUOTE_CLASS(c),&tmp_err);
ali@142	1158	else if (CHAR_IS_SQUOTE(c) && pswit[SQUOTE_SWITCH])
ali@43	1159	{
ali@43	1160	if (s==aline)
ali@43	1161	{
ali@43	1162	/*
ali@142	1163	* At start of line, it can only be a quotation mark.
ali@43	1164	* Hardcode a very common exception!
ali@43	1165	*/
ali@70	1166	if (!g_str_has_prefix(snext,"tis") &&
ali@70	1167	!g_str_has_prefix(snext,"Tis"))
ali@142	1168	(void)count_quote(counters,c,NEUTRAL_QUOTE,&tmp_err);
ali@43	1169	}
ali@70	1170	else if (g_unichar_isalpha(g_utf8_get_char(sprev)) &&
ali@70	1171	g_unichar_isalpha(g_utf8_get_char(snext)))
ali@43	1172	/* Do nothing! it's definitely an apostrophe, not a quote */
ali@43	1173	;
ali@43	1174	/* it's outside a word - let's check it out */
ali@99	1175	else if (c==CHAR_OPEN_SQUOTE \|\| c==CHAR_LS_QUOTE \|\|
ali@70	1176	g_unichar_isalpha(g_utf8_get_char(snext)))
ali@43	1177	{
ali@142	1178	/* certainly looks like a quotation mark */
ali@70	1179	if (!g_str_has_prefix(snext,"tis") &&
ali@70	1180	!g_str_has_prefix(snext,"Tis"))
ali@43	1181	/* hardcode a very common exception! */
ali@142	1182	{
ali@142	1183	if (strchr(".?!,;:",g_utf8_get_char(sprev)))
ali@142	1184	(void)count_quote(counters,c,NEUTRAL_QUOTE,&tmp_err);
ali@142	1185	else
ali@142	1186	(void)count_quote(counters,c,OPENING_QUOTE,&tmp_err);
ali@142	1187	}
ali@43	1188	}
ali@43	1189	else
ali@43	1190	{
ali@142	1191	/* now - is it a quotation mark? */
ali@43	1192	guessquote=0; /* accumulate clues */
ali@70	1193	if (g_unichar_isalpha(g_utf8_get_char(sprev)))
ali@43	1194	{
ali@43	1195	/* it follows a letter - could be either */
ali@43	1196	guessquote++;
ali@70	1197	if (g_utf8_get_char(sprev)=='s')
ali@43	1198	{
ali@43	1199	/* looks like a plural apostrophe */
ali@43	1200	guessquote-=3;
ali@70	1201	if (g_utf8_get_char(snext)==CHAR_SPACE)
ali@70	1202	/* bonus marks! */
ali@43	1203	guessquote-=2;
ali@43	1204	}
ali@142	1205	if (innermost_quote_matches(counters,c))
ali@142	1206	/*
ali@142	1207	* Give it the benefit of some doubt,
ali@142	1208	* if a squote is already open.
ali@142	1209	*/
ali@142	1210	guessquote++;
ali@142	1211	else
ali@142	1212	guessquote--;
ali@142	1213	if (guessquote>=0)
ali@142	1214	(void)count_quote(counters,c,CLOSING_QUOTE,&tmp_err);
ali@43	1215	}
ali@43	1216	else
ali@142	1217	/* no adjacent letter - it must be a quote of some kind */
ali@142	1218	(void)count_quote(counters,c,NEUTRAL_QUOTE,&tmp_err);
ali@43	1219	}
ali@43	1220	}
ali@142	1221	if (tmp_err)
ali@142	1222	{
ali@142	1223	if (pswit[ECHO_SWITCH])
ali@142	1224	g_print("\n%s\n",aline);
ali@142	1225	if (!pswit[OVERVIEW_SWITCH])
ali@142	1226	g_print(" Line %ld column %ld - %s\n",
ali@142	1227	linecnt,g_utf8_pointer_to_offset(aline,s)+1,tmp_err->message);
ali@142	1228	g_clear_error(&tmp_err);
ali@142	1229	}
ali@70	1230	if (c!=CHAR_SPACE && c!='-' && c!='.' && c!=CHAR_ASTERISK &&
ali@70	1231	c!='\r' && c!='\n')
ali@69	1232	isemptyline=FALSE; /* ignore lines like * * * as spacers */
ali@70	1233	if (c==CHAR_UNDERSCORE)
ali@43	1234	counters->c_unders++;
ali@103	1235	if (c==CHAR_OPEN_SBRACK)
ali@103	1236	{
ali@103	1237	if (!matching_difference(counters,COUNTER_ILLUSTRATION) &&
ali@103	1238	!matching_difference(counters,c) && s==aline &&
ali@103	1239	g_str_has_prefix(s,"[Illustration:"))
ali@103	1240	increment_matching(counters,COUNTER_ILLUSTRATION,TRUE);
ali@103	1241	else
ali@103	1242	increment_matching(counters,c,TRUE);
ali@103	1243	}
ali@103	1244	else if (c==CHAR_OPEN_CBRACK \|\| c==CHAR_OPEN_RBRACK)
ali@99	1245	increment_matching(counters,c,TRUE);
ali@103	1246	if (c==CHAR_CLOSE_SBRACK)
ali@103	1247	{
ali@103	1248	if (!matching_count(counters,COUNTER_ILLUSTRATION,FALSE) &&
ali@103	1249	!matching_difference(counters,c) && !*snext)
ali@103	1250	increment_matching(counters,COUNTER_ILLUSTRATION,FALSE);
ali@103	1251	else
ali@103	1252	increment_matching(counters,c,FALSE);
ali@103	1253	}
ali@103	1254	else if (c==CHAR_CLOSE_CBRACK \|\| c==CHAR_CLOSE_RBRACK)
ali@99	1255	increment_matching(counters,c,FALSE);
ali@70	1256	sprev=s;
ali@70	1257	s=snext;
ali@43	1258	}
ali@43	1259	return isemptyline;
ali@43	1260	}
ali@43	1261
ali@41	1262	/*
ali@67	1263	* check_for_control_characters:
ali@67	1264	*
ali@67	1265	* Check for invalid or questionable characters in the line
ali@67	1266	* Anything above 127 is invalid for plain ASCII, and
ali@67	1267	* non-printable control characters should also be flagged.
ali@67	1268	* Tabs should generally not be there.
ali@67	1269	*/
ali@67	1270	void check_for_control_characters(const char *aline)
ali@67	1271	{
ali@70	1272	gunichar c;
ali@67	1273	const char *s;
ali@70	1274	for (s=aline;*s;s=g_utf8_next_char(s))
ali@67	1275	{
ali@70	1276	c=g_utf8_get_char(s);
ali@67	1277	if (c<CHAR_SPACE && c!=CHAR_LF && c!=CHAR_CR && c!=CHAR_TAB)
ali@67	1278	{
ali@67	1279	if (pswit[ECHO_SWITCH])
ali@70	1280	g_print("\n%s\n",aline);
ali@67	1281	if (!pswit[OVERVIEW_SWITCH])
ali@70	1282	g_print(" Line %ld column %ld - Control character %u\n",
ali@70	1283	linecnt,g_utf8_pointer_to_offset(s,aline)+1,c);
ali@67	1284	else
ali@67	1285	cnt_bin++;
ali@67	1286	}
ali@67	1287	}
ali@67	1288	}
ali@67	1289
ali@67	1290	/*
ali@44	1291	* check_for_odd_characters:
ali@44	1292	*
ali@44	1293	* Check for binary and other odd characters.
ali@44	1294	*/
ali@44	1295	void check_for_odd_characters(const char aline,const struct warnings warnings,
ali@69	1296	gboolean isemptyline)
ali@44	1297	{
ali@44	1298	/* Don't repeat multiple warnings on one line. */
ali@185	1299	gboolean eInvalidChar=FALSE,eTab=FALSE,eTilde=FALSE;
ali@70	1300	gboolean eCarat=FALSE,eFSlash=FALSE,eAst=FALSE;
ali@44	1301	const char *s;
ali@70	1302	gunichar c;
ali@185	1303	gsize nb;
ali@185	1304	gchar *t;
ali@70	1305	for (s=aline;*s;s=g_utf8_next_char(s))
ali@44	1306	{
ali@70	1307	c=g_utf8_get_char(s);
ali@185	1308	if (warnings->bin && !eInvalidChar &&
ali@185	1309	(c<CHAR_SPACE && c!='\t' && c!='\n' \|\| c>127))
ali@44	1310	{
ali@44	1311	if (pswit[ECHO_SWITCH])
ali@70	1312	g_print("\n%s\n",aline);
ali@44	1313	if (!pswit[OVERVIEW_SWITCH])
ali@70	1314	if (c>127 && c<160 \|\| c>255)
ali@70	1315	g_print(" Line %ld column %ld - "
ali@70	1316	"Non-ISO-8859 character %u\n",
ali@70	1317	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@44	1318	else
ali@70	1319	g_print(" Line %ld column %ld - "
ali@70	1320	"Non-ASCII character %u\n",
ali@70	1321	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@44	1322	else
ali@44	1323	cnt_bin++;
ali@185	1324	eInvalidChar=TRUE;
ali@185	1325	}
ali@185	1326	if (!eInvalidChar && charset)
ali@185	1327	{
ali@185	1328	if (charset_validator==(GIConv)-1)
ali@185	1329	{
ali@185	1330	if (!g_unichar_isdefined(c))
ali@185	1331	{
ali@185	1332	if (pswit[ECHO_SWITCH])
ali@185	1333	g_print("\n%s\n",aline);
ali@185	1334	if (!pswit[OVERVIEW_SWITCH])
ali@185	1335	g_print(" Line %ld column %ld - Unassigned UNICODE "
ali@185	1336	"code point U+%04" G_GINT32_MODIFIER "X\n",
ali@185	1337	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@185	1338	else
ali@185	1339	cnt_bin++;
ali@185	1340	eInvalidChar=TRUE;
ali@185	1341	}
ali@185	1342	else if (c>=0xE000 && c<=0xF8FF \|\| c>=0xF0000 && c<=0xFFFFD \|\|
ali@185	1343	c>=100000 && c<=0x10FFFD)
ali@185	1344	{
ali@185	1345	if (pswit[ECHO_SWITCH])
ali@185	1346	g_print("\n%s\n",aline);
ali@185	1347	if (!pswit[OVERVIEW_SWITCH])
ali@185	1348	g_print(" Line %ld column %ld - Private Use "
ali@185	1349	"character U+%04" G_GINT32_MODIFIER "X\n",
ali@185	1350	linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
ali@185	1351	else
ali@185	1352	cnt_bin++;
ali@185	1353	eInvalidChar=TRUE;
ali@185	1354	}
ali@185	1355	}
ali@185	1356	else
ali@185	1357	{
ali@185	1358	t=g_convert_with_iconv(s,g_utf8_next_char(s)-s,
ali@185	1359	charset_validator,NULL,&nb,NULL);
ali@185	1360	if (t)
ali@185	1361	g_free(t);
ali@185	1362	else
ali@185	1363	{
ali@185	1364	if (pswit[ECHO_SWITCH])
ali@185	1365	g_print("\n%s\n",aline);
ali@185	1366	if (!pswit[OVERVIEW_SWITCH])
ali@185	1367	g_print(" Line %ld column %ld - Non-%s "
ali@185	1368	"character %u\n",linecnt,
ali@185	1369	g_utf8_pointer_to_offset(aline,s)+1,charset,c);
ali@185	1370	else
ali@185	1371	cnt_bin++;
ali@185	1372	eInvalidChar=TRUE;
ali@185	1373	}
ali@185	1374	}
ali@44	1375	}
ali@70	1376	if (!eTab && c==CHAR_TAB)
ali@44	1377	{
ali@44	1378	if (pswit[ECHO_SWITCH])
ali@70	1379	g_print("\n%s\n",aline);
ali@44	1380	if (!pswit[OVERVIEW_SWITCH])
ali@70	1381	g_print(" Line %ld column %ld - Tab character?\n",
ali@70	1382	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1383	else
ali@44	1384	cnt_odd++;
ali@70	1385	eTab=TRUE;
ali@44	1386	}
ali@70	1387	if (!eTilde && c==CHAR_TILDE)
ali@44	1388	{
ali@44	1389	/*
ali@44	1390	* Often used by OCR software to indicate an
ali@44	1391	* unrecognizable character.
ali@44	1392	*/
ali@44	1393	if (pswit[ECHO_SWITCH])
ali@70	1394	g_print("\n%s\n",aline);
ali@44	1395	if (!pswit[OVERVIEW_SWITCH])
ali@70	1396	g_print(" Line %ld column %ld - Tilde character?\n",
ali@70	1397	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1398	else
ali@44	1399	cnt_odd++;
ali@70	1400	eTilde=TRUE;
ali@44	1401	}
ali@70	1402	if (!eCarat && c==CHAR_CARAT)
ali@44	1403	{
ali@44	1404	if (pswit[ECHO_SWITCH])
ali@70	1405	g_print("\n%s\n",aline);
ali@44	1406	if (!pswit[OVERVIEW_SWITCH])
ali@70	1407	g_print(" Line %ld column %ld - Carat character?\n",
ali@70	1408	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1409	else
ali@44	1410	cnt_odd++;
ali@70	1411	eCarat=TRUE;
ali@44	1412	}
ali@70	1413	if (!eFSlash && c==CHAR_FORESLASH && warnings->fslash)
ali@44	1414	{
ali@44	1415	if (pswit[ECHO_SWITCH])
ali@70	1416	g_print("\n%s\n",aline);
ali@44	1417	if (!pswit[OVERVIEW_SWITCH])
ali@70	1418	g_print(" Line %ld column %ld - Forward slash?\n",
ali@70	1419	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1420	else
ali@44	1421	cnt_odd++;
ali@70	1422	eFSlash=TRUE;
ali@44	1423	}
ali@44	1424	/*
ali@44	1425	* Report asterisks only in paranoid mode,
ali@44	1426	* since they're often deliberate.
ali@44	1427	*/
ali@44	1428	if (!eAst && pswit[PARANOID_SWITCH] && warnings->ast && !isemptyline &&
ali@70	1429	c==CHAR_ASTERISK)
ali@44	1430	{
ali@44	1431	if (pswit[ECHO_SWITCH])
ali@70	1432	g_print("\n%s\n",aline);
ali@44	1433	if (!pswit[OVERVIEW_SWITCH])
ali@70	1434	g_print(" Line %ld column %ld - Asterisk?\n",
ali@70	1435	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@44	1436	else
ali@44	1437	cnt_odd++;
ali@70	1438	eAst=TRUE;
ali@44	1439	}
ali@44	1440	}
ali@44	1441	}
ali@44	1442
ali@44	1443	/*
ali@45	1444	* check_for_long_line:
ali@45	1445	*
ali@45	1446	* Check for line too long.
ali@45	1447	*/
ali@45	1448	void check_for_long_line(const char *aline)
ali@45	1449	{
ali@70	1450	if (g_utf8_strlen(aline,-1)>LONGEST_PG_LINE)
ali@45	1451	{
ali@45	1452	if (pswit[ECHO_SWITCH])
ali@70	1453	g_print("\n%s\n",aline);
ali@45	1454	if (!pswit[OVERVIEW_SWITCH])
ali@70	1455	g_print(" Line %ld column %ld - Long line %ld\n",
ali@70	1456	linecnt,g_utf8_strlen(aline,-1),g_utf8_strlen(aline,-1));
ali@45	1457	else
ali@45	1458	cnt_long++;
ali@45	1459	}
ali@45	1460	}
ali@45	1461
ali@45	1462	/*
ali@45	1463	* check_for_short_line:
ali@45	1464	*
ali@45	1465	* Check for line too short.
ali@45	1466	*
ali@45	1467	* This one is a bit trickier to implement: we don't want to
ali@45	1468	* flag the last line of a paragraph for being short, so we
ali@45	1469	* have to wait until we know that our current line is a
ali@45	1470	* "normal" line, then report the _previous_ line if it was too
ali@45	1471	* short. We also don't want to report indented lines like
ali@45	1472	* chapter heads or formatted quotations. We therefore keep
ali@45	1473	* last->len as the length of the last line examined, and
ali@45	1474	* last->blen as the length of the last but one, and try to
ali@45	1475	* suppress unnecessary warnings by checking that both were of
ali@45	1476	* "normal" length. We keep the first character of the last
ali@45	1477	* line in last->start, and if it was a space, we assume that
ali@45	1478	* the formatting is deliberate. I can't figure out a way to
ali@45	1479	* distinguish something like a quoted verse left-aligned or
ali@45	1480	* the header or footer of a letter from a paragraph of short
ali@45	1481	* lines - maybe if I examined the whole paragraph, and if the
ali@45	1482	* para has less than, say, 8 lines and if all lines are short,
ali@45	1483	* then just assume it's OK? Need to look at some texts to see
ali@45	1484	* how often a formula like this would get the right result.
ali@45	1485	*/
ali@45	1486	void check_for_short_line(const char aline,const struct line_properties last)
ali@45	1487	{
ali@70	1488	if (g_utf8_strlen(aline,-1)>1 && last->len>1 &&
ali@70	1489	last->len<SHORTEST_PG_LINE && last->blen>1 &&
ali@70	1490	last->blen>SHORTEST_PG_LINE && last->start!=CHAR_SPACE)
ali@45	1491	{
ali@45	1492	if (pswit[ECHO_SWITCH])
ali@70	1493	g_print("\n%s\n",prevline);
ali@45	1494	if (!pswit[OVERVIEW_SWITCH])
ali@70	1495	g_print(" Line %ld column %ld - Short line %ld?\n",
ali@70	1496	linecnt-1,g_utf8_strlen(prevline,-1),g_utf8_strlen(prevline,-1));
ali@45	1497	else
ali@45	1498	cnt_short++;
ali@45	1499	}
ali@45	1500	}
ali@45	1501
ali@45	1502	/*
ali@46	1503	* check_for_starting_punctuation:
ali@46	1504	*
ali@46	1505	* Look for punctuation other than full ellipses at start of line.
ali@46	1506	*/
ali@46	1507	void check_for_starting_punctuation(const char *aline)
ali@46	1508	{
ali@70	1509	if (*aline && g_utf8_strchr(".?!,;:",-1,g_utf8_get_char(aline)) &&
ali@70	1510	!g_str_has_prefix(aline,". . ."))
ali@46	1511	{
ali@46	1512	if (pswit[ECHO_SWITCH])
ali@70	1513	g_print("\n%s\n",aline);
ali@46	1514	if (!pswit[OVERVIEW_SWITCH])
ali@70	1515	g_print(" Line %ld column 1 - Begins with punctuation?\n",
ali@46	1516	linecnt);
ali@46	1517	else
ali@46	1518	cnt_punct++;
ali@46	1519	}
ali@46	1520	}
ali@46	1521
ali@46	1522	/*
ali@147	1523	* str_emdash:
ali@147	1524	*
ali@147	1525	* Find the first em-dash, return a pointer to it and set <next> to the
ali@147	1526	* character following the dash.
ali@147	1527	*/
ali@147	1528	char str_emdash(const char s,const char **next)
ali@147	1529	{
ali@147	1530	const char s1,s2;
ali@147	1531	s1=strstr(s,"--");
ali@147	1532	s2=strstr(s,"—");
ali@147	1533	if (!s1)
ali@147	1534	{
ali@147	1535	if (s2)
ali@147	1536	*next=g_utf8_next_char(s2);
ali@147	1537	return (char *)s2;
ali@147	1538	}
ali@147	1539	else if (!s2)
ali@147	1540	{
ali@147	1541	*next=g_utf8_next_char(g_utf8_next_char(s1));
ali@147	1542	return (char *)s1;
ali@147	1543	}
ali@147	1544	else if (s1<s2)
ali@147	1545	{
ali@147	1546	*next=g_utf8_next_char(g_utf8_next_char(s1));
ali@147	1547	return (char *)s1;
ali@147	1548	}
ali@147	1549	else
ali@147	1550	{
ali@147	1551	*next=g_utf8_next_char(s2);
ali@147	1552	return (char *)s2;
ali@147	1553	}
ali@147	1554	}
ali@147	1555
ali@147	1556	/*
ali@47	1557	* check_for_spaced_emdash:
ali@47	1558	*
ali@47	1559	* Check for spaced em-dashes.
ali@47	1560	*
ali@147	1561	* We must check _all_ occurrences of em-dashes on the line
ali@147	1562	* hence the loop - even if the first dash is OK
ali@47	1563	* there may be another that's wrong later on.
ali@47	1564	*/
ali@47	1565	void check_for_spaced_emdash(const char *aline)
ali@47	1566	{
ali@70	1567	const char s,t,*next;
ali@147	1568	for (s=aline;t=str_emdash(s,&next);s=next)
ali@47	1569	{
ali@70	1570	if (t>aline && g_utf8_get_char(g_utf8_prev_char(t))==CHAR_SPACE \|\|
ali@70	1571	g_utf8_get_char(next)==CHAR_SPACE)
ali@47	1572	{
ali@47	1573	if (pswit[ECHO_SWITCH])
ali@70	1574	g_print("\n%s\n",aline);
ali@47	1575	if (!pswit[OVERVIEW_SWITCH])
ali@70	1576	g_print(" Line %ld column %ld - Spaced em-dash?\n",
ali@70	1577	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@47	1578	else
ali@47	1579	cnt_dash++;
ali@47	1580	}
ali@47	1581	}
ali@47	1582	}
ali@47	1583
ali@47	1584	/*
ali@47	1585	* check_for_spaced_dash:
ali@47	1586	*
ali@47	1587	* Check for spaced dashes.
ali@47	1588	*/
ali@47	1589	void check_for_spaced_dash(const char *aline)
ali@47	1590	{
ali@47	1591	const char *s;
ali@47	1592	if ((s=strstr(aline," -")))
ali@47	1593	{
ali@70	1594	if (g_utf8_get_char(g_utf8_next_char(g_utf8_next_char(s)))!='-')
ali@47	1595	{
ali@47	1596	if (pswit[ECHO_SWITCH])
ali@70	1597	g_print("\n%s\n",aline);
ali@47	1598	if (!pswit[OVERVIEW_SWITCH])
ali@70	1599	g_print(" Line %ld column %ld - Spaced dash?\n",
ali@70	1600	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@47	1601	else
ali@47	1602	cnt_dash++;
ali@47	1603	}
ali@47	1604	}
ali@47	1605	else if ((s=strstr(aline,"- ")))
ali@47	1606	{
ali@70	1607	if (s==aline \|\| g_utf8_get_char(g_utf8_prev_char(s))!='-')
ali@47	1608	{
ali@47	1609	if (pswit[ECHO_SWITCH])
ali@70	1610	g_print("\n%s\n",aline);
ali@47	1611	if (!pswit[OVERVIEW_SWITCH])
ali@70	1612	g_print(" Line %ld column %ld - Spaced dash?\n",
ali@70	1613	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@47	1614	else
ali@47	1615	cnt_dash++;
ali@47	1616	}
ali@47	1617	}
ali@47	1618	}
ali@47	1619
ali@47	1620	/*
ali@48	1621	* check_for_unmarked_paragraphs:
ali@48	1622	*
ali@48	1623	* Check for unmarked paragraphs indicated by separate speakers.
ali@48	1624	*
ali@48	1625	* May well be false positive:
ali@48	1626	* "Bravo!" "Wonderful!" called the crowd.
ali@48	1627	* but useful all the same.
ali@48	1628	*/
ali@48	1629	void check_for_unmarked_paragraphs(const char *aline)
ali@48	1630	{
ali@48	1631	const char *s;
ali@48	1632	s=strstr(aline,"\" \"");
ali@48	1633	if (!s)
ali@48	1634	s=strstr(aline,"\" \"");
ali@48	1635	if (s)
ali@48	1636	{
ali@48	1637	if (pswit[ECHO_SWITCH])
ali@70	1638	g_print("\n%s\n",aline);
ali@48	1639	if (!pswit[OVERVIEW_SWITCH])
ali@70	1640	g_print(" Line %ld column %ld - "
ali@70	1641	"Query missing paragraph break?\n",
ali@70	1642	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@48	1643	else
ali@48	1644	cnt_punct++;
ali@48	1645	}
ali@48	1646	}
ali@48	1647
ali@48	1648	/*
ali@49	1649	* check_for_jeebies:
ali@49	1650	*
ali@49	1651	* Check for "to he" and other easy h/b errors.
ali@49	1652	*
ali@49	1653	* This is a very inadequate effort on the h/b problem,
ali@49	1654	* but the phrase "to he" is always an error, whereas "to
ali@49	1655	* be" is quite common.
ali@49	1656	* Similarly, '"Quiet!", be said.' is a non-be error
ali@49	1657	* "to he" is _not_ always an error!:
ali@49	1658	* "Where they went to he couldn't say."
ali@49	1659	* Another false positive:
ali@49	1660	* What would "Cinderella" be without the . . .
ali@49	1661	* and another: "If he wants to he can see for himself."
ali@49	1662	*/
ali@49	1663	void check_for_jeebies(const char *aline)
ali@49	1664	{
ali@49	1665	const char *s;
ali@49	1666	s=strstr(aline," be could ");
ali@49	1667	if (!s)
ali@49	1668	s=strstr(aline," be would ");
ali@49	1669	if (!s)
ali@49	1670	s=strstr(aline," was be ");
ali@49	1671	if (!s)
ali@49	1672	s=strstr(aline," be is ");
ali@49	1673	if (!s)
ali@49	1674	s=strstr(aline," is be ");
ali@49	1675	if (!s)
ali@49	1676	s=strstr(aline,"\", be ");
ali@49	1677	if (!s)
ali@49	1678	s=strstr(aline,"\" be ");
ali@49	1679	if (!s)
ali@49	1680	s=strstr(aline,"\" be ");
ali@49	1681	if (!s)
ali@49	1682	s=strstr(aline," to he ");
ali@49	1683	if (s)
ali@49	1684	{
ali@49	1685	if (pswit[ECHO_SWITCH])
ali@70	1686	g_print("\n%s\n",aline);
ali@49	1687	if (!pswit[OVERVIEW_SWITCH])
ali@70	1688	g_print(" Line %ld column %ld - Query he/be error?\n",
ali@70	1689	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@49	1690	else
ali@49	1691	cnt_word++;
ali@49	1692	}
ali@49	1693	s=strstr(aline," the had ");
ali@49	1694	if (!s)
ali@49	1695	s=strstr(aline," a had ");
ali@49	1696	if (!s)
ali@49	1697	s=strstr(aline," they bad ");
ali@49	1698	if (!s)
ali@49	1699	s=strstr(aline," she bad ");
ali@49	1700	if (!s)
ali@49	1701	s=strstr(aline," he bad ");
ali@49	1702	if (!s)
ali@49	1703	s=strstr(aline," you bad ");
ali@49	1704	if (!s)
ali@49	1705	s=strstr(aline," i bad ");
ali@49	1706	if (s)
ali@49	1707	{
ali@49	1708	if (pswit[ECHO_SWITCH])
ali@70	1709	g_print("\n%s\n",aline);
ali@49	1710	if (!pswit[OVERVIEW_SWITCH])
ali@70	1711	g_print(" Line %ld column %ld - Query had/bad error?\n",
ali@70	1712	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@49	1713	else
ali@49	1714	cnt_word++;
ali@49	1715	}
ali@49	1716	s=strstr(aline,"; hut ");
ali@49	1717	if (!s)
ali@49	1718	s=strstr(aline,", hut ");
ali@49	1719	if (s)
ali@49	1720	{
ali@49	1721	if (pswit[ECHO_SWITCH])
ali@70	1722	g_print("\n%s\n",aline);
ali@49	1723	if (!pswit[OVERVIEW_SWITCH])
ali@70	1724	g_print(" Line %ld column %ld - Query hut/but error?\n",
ali@70	1725	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@49	1726	else
ali@49	1727	cnt_word++;
ali@49	1728	}
ali@49	1729	}
ali@49	1730
ali@49	1731	/*
ali@50	1732	* check_for_mta_from:
ali@50	1733	*
ali@50	1734	* Special case - angled bracket in front of "From" placed there by an
ali@50	1735	* MTA when sending an e-mail.
ali@50	1736	*/
ali@50	1737	void check_for_mta_from(const char *aline)
ali@50	1738	{
ali@50	1739	const char *s;
ali@50	1740	s=strstr(aline,">From");
ali@50	1741	if (s)
ali@50	1742	{
ali@50	1743	if (pswit[ECHO_SWITCH])
ali@70	1744	g_print("\n%s\n",aline);
ali@50	1745	if (!pswit[OVERVIEW_SWITCH])
ali@70	1746	g_print(" Line %ld column %ld - "
ali@70	1747	"Query angled bracket with From\n",
ali@70	1748	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@50	1749	else
ali@50	1750	cnt_punct++;
ali@50	1751	}
ali@50	1752	}
ali@50	1753
ali@50	1754	/*
ali@51	1755	* check_for_orphan_character:
ali@51	1756	*
ali@51	1757	* Check for a single character line -
ali@51	1758	* often an overflow from bad wrapping.
ali@51	1759	*/
ali@51	1760	void check_for_orphan_character(const char *aline)
ali@51	1761	{
ali@70	1762	gunichar c;
ali@70	1763	c=g_utf8_get_char(aline);
ali@70	1764	if (c && !*g_utf8_next_char(aline))
ali@51	1765	{
ali@70	1766	if (c=='I' \|\| c=='V' \|\| c=='X' \|\| c=='L' \|\| g_unichar_isdigit(c))
ali@51	1767	; /* Nothing - ignore numerals alone on a line. */
ali@51	1768	else
ali@51	1769	{
ali@51	1770	if (pswit[ECHO_SWITCH])
ali@70	1771	g_print("\n%s\n",aline);
ali@51	1772	if (!pswit[OVERVIEW_SWITCH])
ali@70	1773	g_print(" Line %ld column 1 - Query single character line\n",
ali@51	1774	linecnt);
ali@51	1775	else
ali@51	1776	cnt_punct++;
ali@51	1777	}
ali@51	1778	}
ali@51	1779	}
ali@51	1780
ali@51	1781	/*
ali@52	1782	* check_for_pling_scanno:
ali@52	1783	*
ali@52	1784	* Check for I" - often should be !
ali@52	1785	*/
ali@52	1786	void check_for_pling_scanno(const char *aline)
ali@52	1787	{
ali@52	1788	const char *s;
ali@52	1789	s=strstr(aline," I\"");
ali@52	1790	if (s)
ali@52	1791	{
ali@52	1792	if (pswit[ECHO_SWITCH])
ali@70	1793	g_print("\n%s\n",aline);
ali@52	1794	if (!pswit[OVERVIEW_SWITCH])
ali@70	1795	g_print(" Line %ld column %ld - Query I=exclamation mark?\n",
ali@70	1796	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@52	1797	else
ali@52	1798	cnt_punct++;
ali@52	1799	}
ali@52	1800	}
ali@52	1801
ali@52	1802	/*
ali@53	1803	* check_for_extra_period:
ali@53	1804	*
ali@53	1805	* Check for period without a capital letter. Cut-down from gutspell.
ali@53	1806	* Only works when it happens on a single line.
ali@53	1807	*/
ali@53	1808	void check_for_extra_period(const char aline,const struct warnings warnings)
ali@53	1809	{
ali@99	1810	const char s,t,s1,sprev;
ali@69	1811	int i;
ali@70	1812	gsize len;
ali@69	1813	gboolean istypo;
ali@69	1814	gchar *testword;
ali@99	1815	gunichar c,nc,pc,*decomposition;
ali@53	1816	if (pswit[PARANOID_SWITCH])
ali@53	1817	{
ali@70	1818	for (t=aline;t=strstr(t,". ");)
ali@53	1819	{
ali@69	1820	if (t==aline)
ali@53	1821	{
ali@70	1822	t=g_utf8_next_char(t);
ali@53	1823	/* start of line punctuation is handled elsewhere */
ali@53	1824	continue;
ali@53	1825	}
ali@70	1826	if (!g_unichar_isalpha(g_utf8_get_char(g_utf8_prev_char(t))))
ali@53	1827	{
ali@70	1828	t=g_utf8_next_char(t);
ali@53	1829	continue;
ali@53	1830	}
ali@53	1831	if (warnings->isDutch)
ali@53	1832	{
ali@53	1833	/* For Frank & Jeroen -- 's Middags case */
ali@70	1834	gunichar c2,c3,c4,c5;
ali@70	1835	c2=g_utf8_get_char(g_utf8_offset_to_pointer(t,2));
ali@70	1836	c3=g_utf8_get_char(g_utf8_offset_to_pointer(t,3));
ali@70	1837	c4=g_utf8_get_char(g_utf8_offset_to_pointer(t,4));
ali@70	1838	c5=g_utf8_get_char(g_utf8_offset_to_pointer(t,5));
ali@99	1839	if (CHAR_IS_APOSTROPHE(c2) &&
ali@99	1840	g_unichar_islower(c3) && c4==CHAR_SPACE &&
ali@99	1841	g_unichar_isupper(c5))
ali@53	1842	{
ali@70	1843	t=g_utf8_next_char(t);
ali@53	1844	continue;
ali@53	1845	}
ali@53	1846	}
ali@70	1847	s1=g_utf8_next_char(g_utf8_next_char(t));
ali@70	1848	while (*s1 && !g_unichar_isalpha(g_utf8_get_char(s1)) &&
ali@173	1849	!g_unichar_isdigit(g_utf8_get_char(s1)))
ali@70	1850	s1=g_utf8_next_char(s1);
ali@70	1851	if (g_unichar_islower(g_utf8_get_char(s1)))
ali@53	1852	{
ali@53	1853	/* we have something to investigate */
ali@69	1854	istypo=TRUE;
ali@53	1855	/* so let's go back and find out */
ali@99	1856	nc=g_utf8_get_char(t);
ali@99	1857	s1=g_utf8_prev_char(t);
ali@99	1858	c=g_utf8_get_char(s1);
ali@99	1859	sprev=g_utf8_prev_char(s1);
ali@99	1860	pc=g_utf8_get_char(sprev);
ali@99	1861	while (s1>=aline &&
ali@99	1862	(g_unichar_isalpha(c) \|\| g_unichar_isdigit(c) \|\|
ali@99	1863	g_unichar_isalpha(pc) && CHAR_IS_APOSTROPHE(c) &&
ali@99	1864	g_unichar_isalpha(nc)))
ali@99	1865	{
ali@99	1866	nc=c;
ali@99	1867	s1=sprev;
ali@99	1868	c=pc;
ali@99	1869	sprev=g_utf8_prev_char(s1);
ali@99	1870	pc=g_utf8_get_char(sprev);
ali@99	1871	}
ali@70	1872	s1=g_utf8_next_char(s1);
ali@69	1873	s=strchr(s1,'.');
ali@69	1874	if (s)
ali@69	1875	testword=g_strndup(s1,s-s1);
ali@69	1876	else
ali@69	1877	testword=g_strdup(s1);
ali@53	1878	for (i=0;*abbrev[i];i++)
ali@53	1879	if (!strcmp(testword,abbrev[i]))
ali@69	1880	istypo=FALSE;
ali@70	1881	if (g_unichar_isdigit(g_utf8_get_char(testword)))
ali@69	1882	istypo=FALSE;
ali@70	1883	if (!*g_utf8_next_char(testword))
ali@69	1884	istypo=FALSE;
ali@53	1885	if (isroman(testword))
ali@69	1886	istypo=FALSE;
ali@53	1887	if (istypo)
ali@53	1888	{
ali@69	1889	istypo=FALSE;
ali@70	1890	for (s=testword;*s;s=g_utf8_next_char(s))
ali@70	1891	{
ali@70	1892	decomposition=g_unicode_canonical_decomposition(
ali@70	1893	g_utf8_get_char(s),&len);
ali@70	1894	if (g_utf8_strchr("aeiou",-1,decomposition[0]))
ali@69	1895	istypo=TRUE;
ali@70	1896	g_free(decomposition);
ali@70	1897	}
ali@53	1898	}
ali@69	1899	if (istypo &&
ali@69	1900	(pswit[VERBOSE_SWITCH] \|\| !g_tree_lookup(qperiod,testword)))
ali@53	1901	{
ali@69	1902	g_tree_insert(qperiod,g_strdup(testword),
ali@69	1903	GINT_TO_POINTER(1));
ali@69	1904	if (pswit[ECHO_SWITCH])
ali@70	1905	g_print("\n%s\n",aline);
ali@69	1906	if (!pswit[OVERVIEW_SWITCH])
ali@70	1907	g_print(" Line %ld column %ld - Extra period?\n",
ali@70	1908	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@69	1909	else
ali@69	1910	cnt_punct++;
ali@53	1911	}
ali@69	1912	g_free(testword);
ali@53	1913	}
ali@70	1914	t=g_utf8_next_char(t);
ali@53	1915	}
ali@53	1916	}
ali@53	1917	}
ali@53	1918
ali@53	1919	/*
ali@54	1920	* check_for_following_punctuation:
ali@54	1921	*
ali@54	1922	* Check for words usually not followed by punctuation.
ali@54	1923	*/
ali@54	1924	void check_for_following_punctuation(const char *aline)
ali@54	1925	{
ali@54	1926	int i;
ali@54	1927	const char s,wordstart;
ali@70	1928	gunichar c;
ali@69	1929	gchar inword,t;
ali@54	1930	if (pswit[TYPO_SWITCH])
ali@54	1931	{
ali@54	1932	for (s=aline;*s;)
ali@54	1933	{
ali@54	1934	wordstart=s;
ali@69	1935	t=getaword(&s);
ali@69	1936	if (!*t)
ali@69	1937	{
ali@69	1938	g_free(t);
ali@54	1939	continue;
ali@69	1940	}
ali@70	1941	inword=g_utf8_strdown(t,-1);
ali@69	1942	g_free(t);
ali@54	1943	for (i=0;*nocomma[i];i++)
ali@54	1944	if (!strcmp(inword,nocomma[i]))
ali@54	1945	{
ali@70	1946	c=g_utf8_get_char(s);
ali@70	1947	if (c==',' \|\| c==';' \|\| c==':')
ali@54	1948	{
ali@54	1949	if (pswit[ECHO_SWITCH])
ali@70	1950	g_print("\n%s\n",aline);
ali@54	1951	if (!pswit[OVERVIEW_SWITCH])
ali@70	1952	g_print(" Line %ld column %ld - "
ali@54	1953	"Query punctuation after %s?\n",
ali@70	1954	linecnt,g_utf8_pointer_to_offset(aline,s)+1,
ali@70	1955	inword);
ali@54	1956	else
ali@54	1957	cnt_punct++;
ali@54	1958	}
ali@54	1959	}
ali@54	1960	for (i=0;*noperiod[i];i++)
ali@54	1961	if (!strcmp(inword,noperiod[i]))
ali@54	1962	{
ali@70	1963	c=g_utf8_get_char(s);
ali@70	1964	if (c=='.' \|\| c=='!')
ali@54	1965	{
ali@54	1966	if (pswit[ECHO_SWITCH])
ali@70	1967	g_print("\n%s\n",aline);
ali@54	1968	if (!pswit[OVERVIEW_SWITCH])
ali@70	1969	g_print(" Line %ld column %ld - "
ali@54	1970	"Query punctuation after %s?\n",
ali@70	1971	linecnt,g_utf8_pointer_to_offset(aline,s)+1,
ali@70	1972	inword);
ali@54	1973	else
ali@54	1974	cnt_punct++;
ali@54	1975	}
ali@54	1976	}
ali@69	1977	g_free(inword);
ali@54	1978	}
ali@54	1979	}
ali@54	1980	}
ali@54	1981
ali@54	1982	/*
ali@55	1983	* check_for_typos:
ali@55	1984	*
ali@55	1985	* Check for commonly mistyped words,
ali@55	1986	* and digits like 0 for O in a word.
ali@55	1987	*/
ali@55	1988	void check_for_typos(const char aline,struct warnings warnings)
ali@55	1989	{
ali@70	1990	const char s,t,nt,wordstart;
ali@70	1991	gchar *inword;
ali@70	1992	gunichar *decomposition;
ali@70	1993	gchar *testword;
ali@70	1994	int i,vowel,consonant,*dupcnt;
ali@70	1995	gboolean isdup,istypo,alower;
ali@99	1996	gunichar c,pc;
ali@70	1997	long offset,len;
ali@70	1998	gsize decomposition_len;
ali@55	1999	for (s=aline;*s;)
ali@55	2000	{
ali@55	2001	wordstart=s;
ali@69	2002	inword=getaword(&s);
ali@55	2003	if (!*inword)
ali@69	2004	{
ali@69	2005	g_free(inword);
ali@55	2006	continue; /* don't bother with empty lines */
ali@69	2007	}
ali@55	2008	if (mixdigit(inword))
ali@55	2009	{
ali@55	2010	if (pswit[ECHO_SWITCH])
ali@70	2011	g_print("\n%s\n",aline);
ali@55	2012	if (!pswit[OVERVIEW_SWITCH])
ali@70	2013	g_print(" Line %ld column %ld - Query digit in %s\n",
ali@70	2014	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+1,inword);
ali@55	2015	else
ali@55	2016	cnt_word++;
ali@55	2017	}
ali@55	2018	/*
ali@55	2019	* Put the word through a series of tests for likely typos and OCR
ali@55	2020	* errors.
ali@55	2021	*/
ali@69	2022	if (pswit[TYPO_SWITCH] \|\| pswit[USERTYPO_SWITCH])
ali@55	2023	{
ali@69	2024	istypo=FALSE;
ali@70	2025	alower=FALSE;
ali@70	2026	for (t=inword;*t;t=g_utf8_next_char(t))
ali@55	2027	{
ali@70	2028	c=g_utf8_get_char(t);
ali@70	2029	nt=g_utf8_next_char(t);
ali@55	2030	/* lowercase for testing */
ali@70	2031	if (g_unichar_islower(c))
ali@70	2032	alower=TRUE;
ali@70	2033	if (alower && (g_unichar_isupper(c) \|\| g_unichar_istitle(c)))
ali@55	2034	{
ali@55	2035	/*
ali@55	2036	* We have an uppercase mid-word. However, there are
ali@55	2037	* common cases:
ali@55	2038	* Mac and Mc like McGill
ali@55	2039	* French contractions like l'Abbe
ali@55	2040	*/
ali@70	2041	offset=g_utf8_pointer_to_offset(inword,t);
ali@99	2042	if (offset>0)
ali@99	2043	pc=g_utf8_get_char(g_utf8_prev_char(t));
ali@99	2044	else
ali@99	2045	pc='\0';
ali@70	2046	if (offset==2 && c=='m' && g_utf8_get_char(nt)=='c' \|\|
ali@70	2047	offset==3 && c=='m' && g_utf8_get_char(nt)=='a' &&
ali@70	2048	g_utf8_get_char(g_utf8_next_char(nt))=='c' \|\|
ali@99	2049	CHAR_IS_APOSTROPHE(pc))
ali@55	2050	; /* do nothing! */
ali@55	2051	else
ali@69	2052	istypo=TRUE;
ali@55	2053	}
ali@55	2054	}
ali@70	2055	testword=g_utf8_casefold(inword,-1);
ali@69	2056	}
ali@69	2057	if (pswit[TYPO_SWITCH])
ali@69	2058	{
ali@55	2059	/*
ali@55	2060	* Check for certain unlikely two-letter combinations at word
ali@55	2061	* start and end.
ali@55	2062	*/
ali@70	2063	len=g_utf8_strlen(testword,-1);
ali@70	2064	if (len>1)
ali@55	2065	{
ali@55	2066	for (i=0;*nostart[i];i++)
ali@70	2067	if (g_str_has_prefix(testword,nostart[i]))
ali@69	2068	istypo=TRUE;
ali@55	2069	for (i=0;*noend[i];i++)
ali@70	2070	if (g_str_has_suffix(testword,noend[i]))
ali@69	2071	istypo=TRUE;
ali@55	2072	}
ali@55	2073	/* ght is common, gbt never. Like that. */
ali@55	2074	if (strstr(testword,"cb"))
ali@69	2075	istypo=TRUE;
ali@55	2076	if (strstr(testword,"gbt"))
ali@69	2077	istypo=TRUE;
ali@55	2078	if (strstr(testword,"pbt"))
ali@69	2079	istypo=TRUE;
ali@55	2080	if (strstr(testword,"tbs"))
ali@69	2081	istypo=TRUE;
ali@55	2082	if (strstr(testword,"mrn"))
ali@69	2083	istypo=TRUE;
ali@55	2084	if (strstr(testword,"ahle"))
ali@69	2085	istypo=TRUE;
ali@55	2086	if (strstr(testword,"ihle"))
ali@69	2087	istypo=TRUE;
ali@55	2088	/*
ali@55	2089	* "TBE" does happen - like HEARTBEAT - but uncommon.
ali@55	2090	* Also "TBI" - frostbite, outbid - but uncommon.
ali@55	2091	* Similarly "ii" like Hawaii, or Pompeii, and in Roman
ali@55	2092	* numerals, but "ii" is a common scanno.
ali@55	2093	*/
ali@55	2094	if (strstr(testword,"tbi"))
ali@69	2095	istypo=TRUE;
ali@55	2096	if (strstr(testword,"tbe"))
ali@69	2097	istypo=TRUE;
ali@55	2098	if (strstr(testword,"ii"))
ali@69	2099	istypo=TRUE;
ali@55	2100	/*
ali@55	2101	* Check for no vowels or no consonants.
ali@55	2102	* If none, flag a typo.
ali@55	2103	*/
ali@70	2104	if (!istypo && len>1)
ali@55	2105	{
ali@55	2106	vowel=consonant=0;
ali@70	2107	for (t=testword;*t;t=g_utf8_next_char(t))
ali@55	2108	{
ali@70	2109	c=g_utf8_get_char(t);
ali@70	2110	decomposition=
ali@70	2111	g_unicode_canonical_decomposition(c,&decomposition_len);
ali@70	2112	if (c=='y' \|\| g_unichar_isdigit(c))
ali@55	2113	{
ali@55	2114	/* Yah, this is loose. */
ali@55	2115	vowel++;
ali@55	2116	consonant++;
ali@55	2117	}
ali@70	2118	else if (g_utf8_strchr("aeiou",-1,decomposition[0]))
ali@55	2119	vowel++;
ali@55	2120	else
ali@55	2121	consonant++;
ali@70	2122	g_free(decomposition);
ali@55	2123	}
ali@55	2124	if (!vowel \|\| !consonant)
ali@69	2125	istypo=TRUE;
ali@55	2126	}
ali@55	2127	/*
ali@55	2128	* Now exclude the word from being reported if it's in
ali@55	2129	* the okword list.
ali@55	2130	*/
ali@55	2131	for (i=0;*okword[i];i++)
ali@55	2132	if (!strcmp(testword,okword[i]))
ali@69	2133	istypo=FALSE;
ali@55	2134	/*
ali@55	2135	* What looks like a typo may be a Roman numeral.
ali@55	2136	* Exclude these.
ali@55	2137	*/
ali@55	2138	if (istypo && isroman(testword))
ali@69	2139	istypo=FALSE;
ali@55	2140	/* Check the manual list of typos. */
ali@55	2141	if (!istypo)
ali@55	2142	for (i=0;*typo[i];i++)
ali@55	2143	if (!strcmp(testword,typo[i]))
ali@69	2144	istypo=TRUE;
ali@55	2145	/*
ali@55	2146	* Check lowercase s, l, i and m - special cases.
ali@55	2147	* "j" - often a semi-colon gone wrong.
ali@55	2148	* "d" for a missing apostrophe - he d
ali@55	2149	* "n" for "in"
ali@55	2150	*/
ali@70	2151	if (!istypo && len==1 &&
ali@70	2152	g_utf8_strchr("slmijdn",-1,g_utf8_get_char(inword)))
ali@69	2153	istypo=TRUE;
ali@55	2154	if (istypo)
ali@55	2155	{
ali@69	2156	dupcnt=g_tree_lookup(qword,testword);
ali@69	2157	if (dupcnt)
ali@69	2158	{
ali@69	2159	(*dupcnt)++;
ali@69	2160	isdup=!pswit[VERBOSE_SWITCH];
ali@69	2161	}
ali@69	2162	else
ali@69	2163	{
ali@69	2164	dupcnt=g_new0(int,1);
ali@69	2165	g_tree_insert(qword,g_strdup(testword),dupcnt);
ali@69	2166	isdup=FALSE;
ali@69	2167	}
ali@55	2168	if (!isdup)
ali@55	2169	{
ali@55	2170	if (pswit[ECHO_SWITCH])
ali@70	2171	g_print("\n%s\n",aline);
ali@55	2172	if (!pswit[OVERVIEW_SWITCH])
ali@55	2173	{
ali@70	2174	g_print(" Line %ld column %ld - Query word %s",
ali@70	2175	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+1,
ali@70	2176	inword);
ali@69	2177	if (!pswit[VERBOSE_SWITCH])
ali@70	2178	g_print(" - not reporting duplicates");
ali@70	2179	g_print("\n");
ali@55	2180	}
ali@55	2181	else
ali@55	2182	cnt_word++;
ali@55	2183	}
ali@55	2184	}
ali@55	2185	}
ali@55	2186	/* check the user's list of typos */
ali@69	2187	if (!istypo && usertypo && g_tree_lookup(usertypo,testword))
ali@69	2188	{
ali@69	2189	if (pswit[ECHO_SWITCH])
ali@70	2190	g_print("\n%s\n",aline);
ali@69	2191	if (!pswit[OVERVIEW_SWITCH])
ali@70	2192	g_print(" Line %ld column %ld - Query possible scanno %s\n",
ali@70	2193	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+2,inword);
ali@69	2194	}
ali@69	2195	if (pswit[TYPO_SWITCH] \|\| pswit[USERTYPO_SWITCH])
ali@69	2196	g_free(testword);
ali@55	2197	if (pswit[PARANOID_SWITCH] && warnings->digit)
ali@55	2198	{
ali@55	2199	/* In paranoid mode, query all 0 and 1 standing alone. */
ali@55	2200	if (!strcmp(inword,"0") \|\| !strcmp(inword,"1"))
ali@55	2201	{
ali@55	2202	if (pswit[ECHO_SWITCH])
ali@70	2203	g_print("\n%s\n",aline);
ali@55	2204	if (!pswit[OVERVIEW_SWITCH])
ali@70	2205	g_print(" Line %ld column %ld - Query standalone %s\n",
ali@70	2206	linecnt,g_utf8_pointer_to_offset(aline,wordstart)+2,
ali@70	2207	inword);
ali@55	2208	else
ali@55	2209	cnt_word++;
ali@55	2210	}
ali@55	2211	}
ali@69	2212	g_free(inword);
ali@55	2213	}
ali@55	2214	}
ali@55	2215
ali@56	2216	/*
ali@56	2217	* check_for_misspaced_punctuation:
ali@56	2218	*
ali@56	2219	* Look for added or missing spaces around punctuation and quotes.
ali@56	2220	* If there is a punctuation character like ! with no space on
ali@56	2221	* either side, suspect a missing!space. If there are spaces on
ali@56	2222	* both sides , assume a typo. If we see a double quote with no
ali@56	2223	* space or punctuation on either side of it, assume unspaced
ali@56	2224	* quotes "like"this.
ali@56	2225	*/
ali@56	2226	void check_for_misspaced_punctuation(const char *aline,
ali@69	2227	struct parities *parities,gboolean isemptyline)
ali@56	2228	{
ali@69	2229	gboolean isacro,isellipsis;
ali@56	2230	const char *s;
ali@70	2231	gunichar c,nc,pc,n2c;
ali@142	2232	int parity;
ali@70	2233	c=g_utf8_get_char(aline);
ali@70	2234	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2235	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2236	{
ali@70	2237	pc=c;
ali@70	2238	c=nc;
ali@70	2239	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2240	/* For each character in the line after the first. */
ali@70	2241	if (g_utf8_strchr(".?!,;:_",-1,c)) /* if it's punctuation */
ali@56	2242	{
ali@56	2243	/* we need to suppress warnings for acronyms like M.D. */
ali@69	2244	isacro=FALSE;
ali@56	2245	/* we need to suppress warnings for ellipsis . . . */
ali@69	2246	isellipsis=FALSE;
ali@70	2247	/*
ali@70	2248	* If there are letters on both sides of it or
ali@70	2249	* if it's strict punctuation followed by an alpha.
ali@70	2250	*/
ali@70	2251	if (g_unichar_isalpha(nc) && (g_unichar_isalpha(pc) \|\|
ali@70	2252	g_utf8_strchr("?!,;:",-1,c)))
ali@56	2253	{
ali@70	2254	if (c=='.')
ali@56	2255	{
ali@70	2256	if (g_utf8_pointer_to_offset(aline,s)>2 &&
ali@70	2257	g_utf8_get_char(g_utf8_offset_to_pointer(s,-2))=='.')
ali@69	2258	isacro=TRUE;
ali@70	2259	n2c=g_utf8_get_char(g_utf8_next_char(g_utf8_next_char(s)));
ali@70	2260	if (nc && n2c=='.')
ali@69	2261	isacro=TRUE;
ali@56	2262	}
ali@56	2263	if (!isacro)
ali@56	2264	{
ali@56	2265	if (pswit[ECHO_SWITCH])
ali@70	2266	g_print("\n%s\n",aline);
ali@56	2267	if (!pswit[OVERVIEW_SWITCH])
ali@70	2268	g_print(" Line %ld column %ld - Missing space?\n",
ali@70	2269	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2270	else
ali@56	2271	cnt_punct++;
ali@56	2272	}
ali@56	2273	}
ali@70	2274	if (pc==CHAR_SPACE && (nc==CHAR_SPACE \|\| !nc))
ali@56	2275	{
ali@56	2276	/*
ali@56	2277	* If there are spaces on both sides,
ali@56	2278	* or space before and end of line.
ali@56	2279	*/
ali@70	2280	if (c=='.')
ali@56	2281	{
ali@70	2282	if (g_utf8_pointer_to_offset(aline,s)>2 &&
ali@70	2283	g_utf8_get_char(g_utf8_offset_to_pointer(s,-2))=='.')
ali@69	2284	isellipsis=TRUE;
ali@70	2285	n2c=g_utf8_get_char(g_utf8_next_char(g_utf8_next_char(s)));
ali@70	2286	if (nc && n2c=='.')
ali@69	2287	isellipsis=TRUE;
ali@56	2288	}
ali@56	2289	if (!isemptyline && !isellipsis)
ali@56	2290	{
ali@56	2291	if (pswit[ECHO_SWITCH])
ali@70	2292	g_print("\n%s\n",aline);
ali@56	2293	if (!pswit[OVERVIEW_SWITCH])
ali@70	2294	g_print(" Line %ld column %ld - "
ali@70	2295	"Spaced punctuation?\n",linecnt,
ali@70	2296	g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2297	else
ali@56	2298	cnt_punct++;
ali@56	2299	}
ali@56	2300	}
ali@56	2301	}
ali@56	2302	}
ali@56	2303	/* Split out the characters that CANNOT be preceded by space. */
ali@70	2304	c=g_utf8_get_char(aline);
ali@70	2305	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2306	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2307	{
ali@70	2308	pc=c;
ali@70	2309	c=nc;
ali@70	2310	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2311	/* for each character in the line after the first */
ali@70	2312	if (g_utf8_strchr("?!,;:",-1,c))
ali@56	2313	{
ali@56	2314	/* if it's punctuation that _cannot_ have a space before it */
ali@70	2315	if (pc==CHAR_SPACE && !isemptyline && nc!=CHAR_SPACE)
ali@56	2316	{
ali@56	2317	/*
ali@70	2318	* If nc DOES == space,
ali@56	2319	* it was already reported just above.
ali@56	2320	*/
ali@56	2321	if (pswit[ECHO_SWITCH])
ali@70	2322	g_print("\n%s\n",aline);
ali@56	2323	if (!pswit[OVERVIEW_SWITCH])
ali@70	2324	g_print(" Line %ld column %ld - Spaced punctuation?\n",
ali@70	2325	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2326	else
ali@56	2327	cnt_punct++;
ali@56	2328	}
ali@56	2329	}
ali@56	2330	}
ali@56	2331	/*
ali@56	2332	* Special case " .X" where X is any alpha.
ali@56	2333	* This plugs a hole in the acronym code above.
ali@56	2334	* Inelegant, but maintainable.
ali@56	2335	*/
ali@70	2336	c=g_utf8_get_char(aline);
ali@70	2337	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2338	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2339	{
ali@70	2340	pc=c;
ali@70	2341	c=nc;
ali@70	2342	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2343	/* for each character in the line after the first */
ali@70	2344	if (c=='.')
ali@56	2345	{
ali@56	2346	/* if it's a period */
ali@70	2347	if (pc==CHAR_SPACE && g_unichar_isalpha(nc))
ali@56	2348	{
ali@56	2349	/*
ali@56	2350	* If the period follows a space and
ali@56	2351	* is followed by a letter.
ali@56	2352	*/
ali@56	2353	if (pswit[ECHO_SWITCH])
ali@70	2354	g_print("\n%s\n",aline);
ali@56	2355	if (!pswit[OVERVIEW_SWITCH])
ali@70	2356	g_print(" Line %ld column %ld - Spaced punctuation?\n",
ali@70	2357	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2358	else
ali@56	2359	cnt_punct++;
ali@56	2360	}
ali@56	2361	}
ali@56	2362	}
ali@70	2363	c=g_utf8_get_char(aline);
ali@70	2364	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2365	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@56	2366	{
ali@70	2367	pc=c;
ali@70	2368	c=nc;
ali@70	2369	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@56	2370	/* for each character in the line after the first */
ali@142	2371	if (CHAR_IS_DQUOTE(c))
ali@56	2372	{
ali@70	2373	if (!g_utf8_strchr(" _-.'`,;:!/([{?}])",-1,pc) &&
ali@70	2374	!g_utf8_strchr(" _-.'`,;:!/([{?}])",-1,nc) && nc \|\|
ali@70	2375	!g_utf8_strchr(" _-([{'`",-1,pc) && g_unichar_isalpha(nc))
ali@56	2376	{
ali@56	2377	if (pswit[ECHO_SWITCH])
ali@70	2378	g_print("\n%s\n",aline);
ali@56	2379	if (!pswit[OVERVIEW_SWITCH])
ali@70	2380	g_print(" Line %ld column %ld - Unspaced quotes?\n",
ali@70	2381	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2382	else
ali@56	2383	cnt_punct++;
ali@56	2384	}
ali@56	2385	}
ali@56	2386	}
ali@56	2387	/* Check parity of quotes. */
ali@70	2388	nc=g_utf8_get_char(aline);
ali@70	2389	for (s=aline;*s;s=g_utf8_next_char(s))
ali@56	2390	{
ali@70	2391	c=nc;
ali@70	2392	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@142	2393	if (CHAR_IS_DQUOTE(c))
ali@56	2394	{
ali@142	2395	if (c==CHAR_DQUOTE)
ali@142	2396	{
ali@142	2397	parities->dquote=!parities->dquote;
ali@142	2398	parity=parities->dquote;
ali@142	2399	}
ali@142	2400	else if (c==CHAR_LD_QUOTE)
ali@142	2401	parity=1;
ali@142	2402	else
ali@142	2403	parity=0;
ali@142	2404	if (!parity)
ali@56	2405	{
ali@56	2406	/* parity even */
ali@173	2407	if (!g_utf8_strchr("_-.'`‘’/,;:!?)]} ",-1,nc))
ali@56	2408	{
ali@56	2409	if (pswit[ECHO_SWITCH])
ali@70	2410	g_print("\n%s\n",aline);
ali@56	2411	if (!pswit[OVERVIEW_SWITCH])
ali@70	2412	g_print(" Line %ld column %ld - "
ali@70	2413	"Wrongspaced quotes?\n",
ali@70	2414	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2415	else
ali@56	2416	cnt_punct++;
ali@56	2417	}
ali@56	2418	}
ali@56	2419	else
ali@56	2420	{
ali@56	2421	/* parity odd */
ali@173	2422	if (!g_unichar_isalpha(nc) && !g_unichar_isdigit(nc) &&
ali@173	2423	!g_utf8_strchr("_-/.'`‘’([{$",-1,nc) \|\| !nc)
ali@56	2424	{
ali@56	2425	if (pswit[ECHO_SWITCH])
ali@70	2426	g_print("\n%s\n",aline);
ali@56	2427	if (!pswit[OVERVIEW_SWITCH])
ali@70	2428	g_print(" Line %ld column %ld - "
ali@70	2429	"Wrongspaced quotes?\n",
ali@70	2430	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2431	else
ali@56	2432	cnt_punct++;
ali@56	2433	}
ali@56	2434	}
ali@56	2435	}
ali@56	2436	}
ali@142	2437	c=g_utf8_get_char(aline);
ali@142	2438	if (CHAR_IS_DQUOTE(c))
ali@56	2439	{
ali@70	2440	if (g_utf8_strchr(",;:!?)]} ",-1,
ali@70	2441	g_utf8_get_char(g_utf8_next_char(aline))))
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 1 - Wrongspaced quotes?\n",
ali@56	2447	linecnt);
ali@56	2448	else
ali@56	2449	cnt_punct++;
ali@56	2450	}
ali@56	2451	}
ali@56	2452	if (pswit[SQUOTE_SWITCH])
ali@56	2453	{
ali@70	2454	nc=g_utf8_get_char(aline);
ali@70	2455	for (s=aline;*s;s=g_utf8_next_char(s))
ali@56	2456	{
ali@70	2457	c=nc;
ali@70	2458	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@99	2459	if (CHAR_IS_SQUOTE(c) && (s==aline \|\| s>aline &&
ali@70	2460	!g_unichar_isalpha(g_utf8_get_char(g_utf8_prev_char(s))) \|\|
ali@70	2461	!g_unichar_isalpha(nc)))
ali@56	2462	{
ali@56	2463	parities->squote=!parities->squote;
ali@56	2464	if (!parities->squote)
ali@56	2465	{
ali@56	2466	/* parity even */
ali@70	2467	if (!g_utf8_strchr("_-.'`/\",;:!?)]} ",-1,nc))
ali@56	2468	{
ali@56	2469	if (pswit[ECHO_SWITCH])
ali@70	2470	g_print("\n%s\n",aline);
ali@56	2471	if (!pswit[OVERVIEW_SWITCH])
ali@70	2472	g_print(" Line %ld column %ld - "
ali@56	2473	"Wrongspaced singlequotes?\n",
ali@70	2474	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2475	else
ali@56	2476	cnt_punct++;
ali@56	2477	}
ali@56	2478	}
ali@56	2479	else
ali@56	2480	{
ali@56	2481	/* parity odd */
ali@173	2482	if (!g_unichar_isalpha(nc) && !g_unichar_isdigit(nc) &&
ali@70	2483	!g_utf8_strchr("_-/\".'`",-1,nc) \|\| !nc)
ali@56	2484	{
ali@56	2485	if (pswit[ECHO_SWITCH])
ali@70	2486	g_print("\n%s\n",aline);
ali@56	2487	if (!pswit[OVERVIEW_SWITCH])
ali@70	2488	g_print(" Line %ld column %ld - "
ali@56	2489	"Wrongspaced singlequotes?\n",
ali@70	2490	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@56	2491	else
ali@56	2492	cnt_punct++;
ali@56	2493	}
ali@56	2494	}
ali@56	2495	}
ali@56	2496	}
ali@56	2497	}
ali@56	2498	}
ali@56	2499
ali@55	2500	/*
ali@57	2501	* check_for_double_punctuation:
ali@57	2502	*
ali@57	2503	* Look for double punctuation like ,. or ,,
ali@57	2504	* Thanks to DW for the suggestion!
ali@57	2505	* In books with references, ".," and ".;" are common
ali@57	2506	* e.g. "etc., etc.," and vol. 1.; vol 3.;
ali@57	2507	* OTOH, from my initial tests, there are also fairly
ali@57	2508	* common errors. What to do? Make these cases paranoid?
ali@57	2509	* ".," is the most common, so warnings->dotcomma is used
ali@57	2510	* to suppress detailed reporting if it occurs often.
ali@57	2511	*/
ali@57	2512	void check_for_double_punctuation(const char aline,struct warnings warnings)
ali@57	2513	{
ali@70	2514	const char *s;
ali@70	2515	gunichar c,nc;
ali@70	2516	nc=g_utf8_get_char(aline);
ali@70	2517	for (s=aline;*s;s=g_utf8_next_char(s))
ali@57	2518	{
ali@70	2519	c=nc;
ali@70	2520	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@57	2521	/* for each punctuation character in the line */
ali@70	2522	if (c && nc && g_utf8_strchr(".?!,;:",-1,c) &&
ali@70	2523	g_utf8_strchr(".?!,;:",-1,nc))
ali@57	2524	{
ali@57	2525	/* followed by punctuation, it's a query, unless . . . */
ali@70	2526	if (c==nc && (c=='.' \|\| c=='?' \|\| c=='!') \|\|
ali@70	2527	!warnings->dotcomma && c=='.' && nc==',' \|\|
ali@70	2528	warnings->isFrench && g_str_has_prefix(s,",...") \|\|
ali@70	2529	warnings->isFrench && g_str_has_prefix(s,"...,") \|\|
ali@70	2530	warnings->isFrench && g_str_has_prefix(s,";...") \|\|
ali@70	2531	warnings->isFrench && g_str_has_prefix(s,"...;") \|\|
ali@70	2532	warnings->isFrench && g_str_has_prefix(s,":...") \|\|
ali@70	2533	warnings->isFrench && g_str_has_prefix(s,"...:") \|\|
ali@70	2534	warnings->isFrench && g_str_has_prefix(s,"!...") \|\|
ali@70	2535	warnings->isFrench && g_str_has_prefix(s,"...!") \|\|
ali@70	2536	warnings->isFrench && g_str_has_prefix(s,"?...") \|\|
ali@70	2537	warnings->isFrench && g_str_has_prefix(s,"...?"))
ali@57	2538	{
ali@70	2539	if (warnings->isFrench && g_str_has_prefix(s,",...") \|\|
ali@70	2540	warnings->isFrench && g_str_has_prefix(s,"...,") \|\|
ali@70	2541	warnings->isFrench && g_str_has_prefix(s,";...") \|\|
ali@70	2542	warnings->isFrench && g_str_has_prefix(s,"...;") \|\|
ali@70	2543	warnings->isFrench && g_str_has_prefix(s,":...") \|\|
ali@70	2544	warnings->isFrench && g_str_has_prefix(s,"...:") \|\|
ali@70	2545	warnings->isFrench && g_str_has_prefix(s,"!...") \|\|
ali@70	2546	warnings->isFrench && g_str_has_prefix(s,"...!") \|\|
ali@70	2547	warnings->isFrench && g_str_has_prefix(s,"?...") \|\|
ali@70	2548	warnings->isFrench && g_str_has_prefix(s,"...?"))
ali@70	2549	{
ali@70	2550	s+=4;
ali@70	2551	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@70	2552	}
ali@57	2553	; /* do nothing for .. !! and ?? which can be legit */
ali@57	2554	}
ali@57	2555	else
ali@57	2556	{
ali@57	2557	if (pswit[ECHO_SWITCH])
ali@70	2558	g_print("\n%s\n",aline);
ali@57	2559	if (!pswit[OVERVIEW_SWITCH])
ali@70	2560	g_print(" Line %ld column %ld - Double punctuation?\n",
ali@70	2561	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@57	2562	else
ali@57	2563	cnt_punct++;
ali@57	2564	}
ali@57	2565	}
ali@57	2566	}
ali@57	2567	}
ali@57	2568
ali@57	2569	/*
ali@58	2570	* check_for_spaced_quotes:
ali@58	2571	*/
ali@58	2572	void check_for_spaced_quotes(const char *aline)
ali@58	2573	{
ali@99	2574	int i;
ali@58	2575	const char s,t;
ali@99	2576	const gunichar single_quotes[]={CHAR_SQUOTE,CHAR_OPEN_SQUOTE,CHAR_LS_QUOTE,
ali@99	2577	CHAR_RS_QUOTE};
ali@99	2578	GString *pattern;
ali@58	2579	s=aline;
ali@58	2580	while ((t=strstr(s," \" ")))
ali@58	2581	{
ali@58	2582	if (pswit[ECHO_SWITCH])
ali@70	2583	g_print("\n%s\n",aline);
ali@58	2584	if (!pswit[OVERVIEW_SWITCH])
ali@70	2585	g_print(" Line %ld column %ld - Spaced doublequote?\n",
ali@70	2586	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@58	2587	else
ali@58	2588	cnt_punct++;
ali@70	2589	s=g_utf8_next_char(g_utf8_next_char(t));
ali@58	2590	}
ali@99	2591	pattern=g_string_new(NULL);
ali@99	2592	for(i=0;i<G_N_ELEMENTS(single_quotes);i++)
ali@58	2593	{
ali@99	2594	g_string_assign(pattern," ");
ali@99	2595	g_string_append_unichar(pattern,single_quotes[i]);
ali@99	2596	g_string_append_c(pattern,' ');
ali@99	2597	s=aline;
ali@99	2598	while ((t=strstr(s,pattern->str)))
ali@99	2599	{
ali@99	2600	if (pswit[ECHO_SWITCH])
ali@99	2601	g_print("\n%s\n",aline);
ali@99	2602	if (!pswit[OVERVIEW_SWITCH])
ali@99	2603	g_print(" Line %ld column %ld - Spaced singlequote?\n",
ali@99	2604	linecnt,g_utf8_pointer_to_offset(aline,t)+1);
ali@99	2605	else
ali@99	2606	cnt_punct++;
ali@99	2607	s=g_utf8_next_char(g_utf8_next_char(t));
ali@99	2608	}
ali@58	2609	}
ali@99	2610	g_string_free(pattern,TRUE);
ali@58	2611	}
ali@58	2612
ali@58	2613	/*
ali@59	2614	* check_for_miscased_genative:
ali@59	2615	*
ali@59	2616	* Check special case of 'S instead of 's at end of word.
ali@59	2617	*/
ali@59	2618	void check_for_miscased_genative(const char *aline)
ali@59	2619	{
ali@59	2620	const char *s;
ali@70	2621	gunichar c,nc,pc;
ali@69	2622	if (!*aline)
ali@69	2623	return;
ali@70	2624	c=g_utf8_get_char(aline);
ali@70	2625	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2626	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@59	2627	{
ali@70	2628	pc=c;
ali@70	2629	c=nc;
ali@70	2630	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@99	2631	if (CHAR_IS_APOSTROPHE(c) && nc=='S' && g_unichar_islower(pc))
ali@59	2632	{
ali@59	2633	if (pswit[ECHO_SWITCH])
ali@70	2634	g_print("\n%s\n",aline);
ali@59	2635	if (!pswit[OVERVIEW_SWITCH])
ali@70	2636	g_print(" Line %ld column %ld - Capital \"S\"?\n",
ali@70	2637	linecnt,g_utf8_pointer_to_offset(aline,s)+2);
ali@59	2638	else
ali@59	2639	cnt_punct++;
ali@59	2640	}
ali@59	2641	}
ali@59	2642	}
ali@59	2643
ali@59	2644	/*
ali@60	2645	* check_end_of_line:
ali@60	2646	*
ali@60	2647	* Now check special cases - start and end of line -
ali@60	2648	* for single and double quotes. Start is sometimes [sic]
ali@60	2649	* but better to query it anyway.
ali@60	2650	* While we're here, check for dash at end of line.
ali@60	2651	*/
ali@60	2652	void check_end_of_line(const char aline,struct warnings warnings)
ali@60	2653	{
ali@70	2654	int lbytes;
ali@70	2655	const char *s;
ali@70	2656	gunichar c1,c2;
ali@70	2657	lbytes=strlen(aline);
ali@70	2658	if (g_utf8_strlen(aline,lbytes)>1)
ali@60	2659	{
ali@70	2660	s=g_utf8_prev_char(aline+lbytes);
ali@70	2661	c1=g_utf8_get_char(s);
ali@70	2662	c2=g_utf8_get_char(g_utf8_prev_char(s));
ali@142	2663	if ((CHAR_IS_DQUOTE(c1) \|\| CHAR_IS_SQUOTE(c1)) && c2==CHAR_SPACE)
ali@60	2664	{
ali@60	2665	if (pswit[ECHO_SWITCH])
ali@70	2666	g_print("\n%s\n",aline);
ali@60	2667	if (!pswit[OVERVIEW_SWITCH])
ali@70	2668	g_print(" Line %ld column %ld - Spaced quote?\n",linecnt,
ali@70	2669	g_utf8_strlen(aline,lbytes));
ali@70	2670	else
ali@70	2671	cnt_punct++;
ali@70	2672	}
ali@70	2673	c1=g_utf8_get_char(aline);
ali@70	2674	c2=g_utf8_get_char(g_utf8_next_char(aline));
ali@99	2675	if (CHAR_IS_SQUOTE(c1) && c2==CHAR_SPACE)
ali@70	2676	{
ali@70	2677	if (pswit[ECHO_SWITCH])
ali@70	2678	g_print("\n%s\n",aline);
ali@70	2679	if (!pswit[OVERVIEW_SWITCH])
ali@70	2680	g_print(" Line %ld column 1 - Spaced quote?\n",linecnt);
ali@60	2681	else
ali@60	2682	cnt_punct++;
ali@60	2683	}
ali@60	2684	/*
ali@60	2685	* Dash at end of line may well be legit - paranoid mode only
ali@60	2686	* and don't report em-dash at line-end.
ali@60	2687	*/
ali@60	2688	if (pswit[PARANOID_SWITCH] && warnings->hyphen)
ali@60	2689	{
ali@70	2690	for (s=g_utf8_prev_char(aline+lbytes);
ali@70	2691	s>aline && g_utf8_get_char(s)<=CHAR_SPACE;s=g_utf8_prev_char(s))
ali@60	2692	;
ali@70	2693	if (g_utf8_get_char(s)=='-' &&
ali@70	2694	g_utf8_get_char(g_utf8_prev_char(s))!='-')
ali@60	2695	{
ali@60	2696	if (pswit[ECHO_SWITCH])
ali@70	2697	g_print("\n%s\n",aline);
ali@60	2698	if (!pswit[OVERVIEW_SWITCH])
ali@70	2699	g_print(" Line %ld column %ld - "
ali@70	2700	"Hyphen at end of line?\n",
ali@70	2701	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@60	2702	}
ali@60	2703	}
ali@60	2704	}
ali@60	2705	}
ali@60	2706
ali@60	2707	/*
ali@61	2708	* check_for_unspaced_bracket:
ali@61	2709	*
ali@61	2710	* Brackets are often unspaced, but shouldn't be surrounded by alpha.
ali@61	2711	* If so, suspect a scanno like "a]most".
ali@61	2712	*/
ali@61	2713	void check_for_unspaced_bracket(const char *aline)
ali@61	2714	{
ali@70	2715	const char *s;
ali@70	2716	gunichar c,nc,pc;
ali@70	2717	c=g_utf8_get_char(aline);
ali@70	2718	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2719	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@61	2720	{
ali@70	2721	pc=c;
ali@70	2722	c=nc;
ali@70	2723	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@70	2724	if (!nc)
ali@70	2725	break;
ali@61	2726	/* for each bracket character in the line except 1st & last */
ali@70	2727	if (g_utf8_strchr("{[()]}",-1,c) &&
ali@70	2728	g_unichar_isalpha(pc) && g_unichar_isalpha(nc))
ali@61	2729	{
ali@61	2730	if (pswit[ECHO_SWITCH])
ali@70	2731	g_print("\n%s\n",aline);
ali@61	2732	if (!pswit[OVERVIEW_SWITCH])
ali@70	2733	g_print(" Line %ld column %ld - Unspaced bracket?\n",
ali@70	2734	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@61	2735	else
ali@61	2736	cnt_punct++;
ali@61	2737	}
ali@61	2738	}
ali@61	2739	}
ali@61	2740
ali@61	2741	/*
ali@62	2742	* check_for_unpunctuated_endquote:
ali@62	2743	*/
ali@62	2744	void check_for_unpunctuated_endquote(const char *aline)
ali@62	2745	{
ali@70	2746	const char *s;
ali@70	2747	gunichar c,nc,pc;
ali@142	2748	QuoteClass qc;
ali@70	2749	c=g_utf8_get_char(aline);
ali@70	2750	nc=c?g_utf8_get_char(g_utf8_next_char(aline)):0;
ali@70	2751	for (s=g_utf8_next_char(aline);nc;s=g_utf8_next_char(s))
ali@62	2752	{
ali@70	2753	pc=c;
ali@70	2754	c=nc;
ali@142	2755	qc=CHAR_IS_DQUOTE(c)?QUOTE_CLASS(c):INVALID_QUOTE;
ali@70	2756	nc=g_utf8_get_char(g_utf8_next_char(s));
ali@62	2757	/* for each character in the line except 1st */
ali@147	2758	if ((qc==CLOSING_QUOTE \|\| qc==NEUTRAL_QUOTE) && g_unichar_isalpha(pc))
ali@62	2759	{
ali@62	2760	if (pswit[ECHO_SWITCH])
ali@70	2761	g_print("\n%s\n",aline);
ali@62	2762	if (!pswit[OVERVIEW_SWITCH])
ali@70	2763	g_print(" Line %ld column %ld - "
ali@70	2764	"endquote missing punctuation?\n",
ali@70	2765	linecnt,g_utf8_pointer_to_offset(aline,s));
ali@62	2766	else
ali@62	2767	cnt_punct++;
ali@62	2768	}
ali@62	2769	}
ali@62	2770	}
ali@62	2771
ali@62	2772	/*
ali@63	2773	* check_for_html_tag:
ali@63	2774	*
ali@63	2775	* Check for <HTML TAG>.
ali@63	2776	*
ali@63	2777	* If there is a < in the line, followed at some point
ali@63	2778	* by a > then we suspect HTML.
ali@63	2779	*/
ali@63	2780	void check_for_html_tag(const char *aline)
ali@63	2781	{
ali@63	2782	const char open,close;
ali@70	2783	gchar *tag;
ali@70	2784	open=strchr(aline,'<');
ali@63	2785	if (open)
ali@63	2786	{
ali@70	2787	close=strchr(g_utf8_next_char(open),'>');
ali@63	2788	if (close)
ali@63	2789	{
ali@70	2790	if (pswit[ECHO_SWITCH])
ali@70	2791	g_print("\n%s\n",aline);
ali@70	2792	if (!pswit[OVERVIEW_SWITCH])
ali@63	2793	{
ali@70	2794	tag=g_strndup(open,close-open+1);
ali@70	2795	g_print(" Line %ld column %ld - HTML Tag? %s \n",
ali@70	2796	linecnt,g_utf8_pointer_to_offset(aline,open)+1,tag);
ali@70	2797	g_free(tag);
ali@63	2798	}
ali@70	2799	else
ali@70	2800	cnt_html++;
ali@63	2801	}
ali@63	2802	}
ali@63	2803	}
ali@63	2804
ali@63	2805	/*
ali@64	2806	* check_for_html_entity:
ali@64	2807	*
ali@64	2808	* Check for &symbol; HTML.
ali@64	2809	*
ali@64	2810	* If there is a & in the line, followed at
ali@64	2811	* some point by a ; then we suspect HTML.
ali@64	2812	*/
ali@64	2813	void check_for_html_entity(const char *aline)
ali@64	2814	{
ali@64	2815	const char s,amp,*scolon;
ali@70	2816	gchar *entity;
ali@70	2817	amp=strchr(aline,'&');
ali@64	2818	if (amp)
ali@64	2819	{
ali@70	2820	scolon=strchr(amp,';');
ali@64	2821	if (scolon)
ali@64	2822	{
ali@70	2823	for (s=amp;s<scolon;s=g_utf8_next_char(s))
ali@70	2824	if (g_utf8_get_char(s)==CHAR_SPACE)
ali@70	2825	break; /* Don't report "Jones & Son;" */
ali@70	2826	if (s>=scolon)
ali@64	2827	{
ali@64	2828	if (pswit[ECHO_SWITCH])
ali@70	2829	g_print("\n%s\n",aline);
ali@64	2830	if (!pswit[OVERVIEW_SWITCH])
ali@70	2831	{
ali@70	2832	entity=g_strndup(amp,scolon-amp+1);
ali@70	2833	g_print(" Line %ld column %d - HTML symbol? %s \n",
ali@70	2834	linecnt,(int)(amp-aline)+1,entity);
ali@70	2835	g_free(entity);
ali@70	2836	}
ali@64	2837	else
ali@64	2838	cnt_html++;
ali@64	2839	}
ali@64	2840	}
ali@64	2841	}
ali@64	2842	}
ali@64	2843
ali@65	2844	/*
ali@66	2845	* check_for_omitted_punctuation:
ali@66	2846	*
ali@66	2847	* Check for omitted punctuation at end of paragraph by working back
ali@66	2848	* through prevline. DW.
ali@66	2849	* Need to check this only for "normal" paras.
ali@66	2850	* So what is a "normal" para?
ali@66	2851	* Not normal if one-liner (chapter headings, etc.)
ali@66	2852	* Not normal if doesn't contain at least one locase letter
ali@66	2853	* Not normal if starts with space
ali@66	2854	*/
ali@66	2855	void check_for_omitted_punctuation(const char *prevline,
ali@66	2856	struct line_properties *last,int start_para_line)
ali@66	2857	{
ali@70	2858	gboolean letter_on_line=FALSE;
ali@66	2859	const char *s;
ali@99	2860	gunichar c;
ali@142	2861	gboolean closing_quote;
ali@70	2862	for (s=prevline;*s;s=g_utf8_next_char(s))
ali@70	2863	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@70	2864	{
ali@70	2865	letter_on_line=TRUE;
ali@70	2866	break;
ali@70	2867	}
ali@66	2868	/*
ali@66	2869	* This next "if" is a problem.
ali@66	2870	* If we say "start_para_line <= linecnt - 1", that includes
ali@66	2871	* one-line "paragraphs" like chapter heads. Lotsa false positives.
ali@66	2872	* If we say "start_para_line < linecnt - 1" it doesn't, but then it
ali@66	2873	* misses genuine one-line paragraphs.
ali@66	2874	*/
ali@70	2875	if (letter_on_line && last->blen>2 && start_para_line<linecnt-1 &&
ali@70	2876	g_utf8_get_char(prevline)>CHAR_SPACE)
ali@66	2877	{
ali@99	2878	s=prevline+strlen(prevline);
ali@99	2879	do
ali@99	2880	{
ali@99	2881	s=g_utf8_prev_char(s);
ali@99	2882	c=g_utf8_get_char(s);
ali@142	2883	if (QUOTE_CLASS(c)==CLOSING_QUOTE \|\| QUOTE_CLASS(c)==NEUTRAL_QUOTE)
ali@142	2884	closing_quote=TRUE;
ali@142	2885	else
ali@142	2886	closing_quote=FALSE;
ali@142	2887	} while (closing_quote && s>prevline);
ali@70	2888	for (;s>prevline;s=g_utf8_prev_char(s))
ali@66	2889	{
ali@70	2890	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@66	2891	{
ali@66	2892	if (pswit[ECHO_SWITCH])
ali@70	2893	g_print("\n%s\n",prevline);
ali@66	2894	if (!pswit[OVERVIEW_SWITCH])
ali@70	2895	g_print(" Line %ld column %ld - "
ali@66	2896	"No punctuation at para end?\n",
ali@70	2897	linecnt-1,g_utf8_strlen(prevline,-1));
ali@66	2898	else
ali@66	2899	cnt_punct++;
ali@66	2900	break;
ali@66	2901	}
ali@147	2902	if (g_utf8_strchr("—-.:!([{?}])",-1,g_utf8_get_char(s)))
ali@66	2903	break;
ali@66	2904	}
ali@66	2905	}
ali@66	2906	}
ali@66	2907
ali@69	2908	gboolean report_duplicate_queries(gpointer key,gpointer value,gpointer data)
ali@69	2909	{
ali@69	2910	const char *word=key;
ali@69	2911	int *dupcnt=value;
ali@69	2912	if (*dupcnt)
ali@70	2913	g_print("\nNote: Queried word %s was duplicated %d times\n",
ali@69	2914	word,*dupcnt);
ali@69	2915	return FALSE;
ali@69	2916	}
ali@69	2917
ali@70	2918	void print_as_windows_1252(const char *string)
ali@70	2919	{
ali@70	2920	gsize inbytes,outbytes;
ali@70	2921	gchar buf,bp;
ali@86	2922	static GIConv converter=(GIConv)-1;
ali@70	2923	if (!string)
ali@70	2924	{
ali@70	2925	if (converter!=(GIConv)-1)
ali@70	2926	g_iconv_close(converter);
ali@70	2927	converter=(GIConv)-1;
ali@70	2928	return;
ali@70	2929	}
ali@86	2930	if (converter==(GIConv)-1)
ali@70	2931	converter=g_iconv_open("WINDOWS-1252","UTF-8");
ali@70	2932	if (converter!=(GIConv)-1)
ali@70	2933	{
ali@70	2934	inbytes=outbytes=strlen(string);
ali@70	2935	bp=buf=g_malloc(outbytes+1);
ali@70	2936	g_iconv(converter,(char **)&string,&inbytes,&bp,&outbytes);
ali@70	2937	*bp='\0';
ali@70	2938	fputs(buf,stdout);
ali@70	2939	g_free(buf);
ali@70	2940	}
ali@70	2941	else
ali@70	2942	fputs(string,stdout);
ali@70	2943	}
ali@70	2944
ali@72	2945	void print_as_utf_8(const char *string)
ali@72	2946	{
ali@72	2947	fputs(string,stdout);
ali@72	2948	}
ali@72	2949
ali@66	2950	/*
ali@41	2951	* procfile:
ali@41	2952	*
ali@41	2953	* Process one file.
ali@41	2954	*/
ali@69	2955	void procfile(const char *filename)
ali@41	2956	{
ali@65	2957	const char *s;
ali@69	2958	gchar parastart=NULL; / first line of current para */
ali@69	2959	gchar etext,aline;
ali@69	2960	gchar *etext_ptr;
ali@69	2961	GError *err=NULL;
ali@41	2962	struct first_pass_results *first_pass_results;
ali@42	2963	struct warnings *warnings;
ali@43	2964	struct counters counters={0};
ali@45	2965	struct line_properties last={0};
ali@56	2966	struct parities parities={0};
ali@69	2967	struct pending pending={0};
ali@69	2968	gboolean isemptyline;
ali@68	2969	long start_para_line=0;
ali@69	2970	gboolean isnewpara=FALSE,enddash=FALSE;
ali@45	2971	last.start=CHAR_SPACE;
ali@68	2972	linecnt=checked_linecnt=0;
ali@69	2973	etext=read_etext(filename,&err);
ali@69	2974	if (!etext)
ali@41	2975	{
ali@68	2976	if (pswit[STDOUT_SWITCH])
ali@69	2977	fprintf(stdout,"bookloupe: %s: %s\n",filename,err->message);
ali@68	2978	else
ali@69	2979	fprintf(stderr,"bookloupe: %s: %s\n",filename,err->message);
ali@41	2980	exit(1);
ali@41	2981	}
ali@70	2982	g_print("\n\nFile: %s\n\n",filename);
ali@69	2983	first_pass_results=first_pass(etext);
ali@42	2984	warnings=report_first_pass(first_pass_results);
ali@69	2985	qword=g_tree_new_full((GCompareDataFunc)strcmp,NULL,g_free,g_free);
ali@69	2986	qperiod=g_tree_new_full((GCompareDataFunc)strcmp,NULL,g_free,NULL);
ali@40	2987	/*
ali@40	2988	* Here we go with the main pass. Hold onto yer hat!
ali@40	2989	*/
ali@65	2990	linecnt=0;
ali@69	2991	etext_ptr=etext;
ali@69	2992	while ((aline=flgets(&etext_ptr,linecnt+1)))
ali@40	2993	{
ali@68	2994	linecnt++;
ali@68	2995	if (linecnt==1)
ali@69	2996	isnewpara=TRUE;
ali@70	2997	if (pswit[DP_SWITCH] && g_str_has_prefix(aline,"-----File: "))
ali@40	2998	continue; // skip DP page separators completely
ali@68	2999	if (linecnt<first_pass_results->firstline \|\|
ali@41	3000	(first_pass_results->footerline>0 &&
ali@41	3001	linecnt>first_pass_results->footerline))
ali@40	3002	{
ali@68	3003	if (pswit[HEADER_SWITCH])
ali@40	3004	{
ali@70	3005	if (g_str_has_prefix(aline,"Title:"))
ali@70	3006	g_print(" %s\n",aline);
ali@70	3007	if (g_str_has_prefix(aline,"Author:"))
ali@70	3008	g_print(" %s\n",aline);
ali@70	3009	if (g_str_has_prefix(aline,"Release Date:"))
ali@70	3010	g_print(" %s\n",aline);
ali@70	3011	if (g_str_has_prefix(aline,"Edition:"))
ali@70	3012	g_print(" %s\n\n",aline);
ali@40	3013	}
ali@68	3014	continue; /* skip through the header */
ali@40	3015	}
ali@68	3016	checked_linecnt++;
ali@65	3017	print_pending(aline,parastart,&pending);
ali@164	3018	isemptyline=analyse_quotes(aline,&counters);
ali@68	3019	if (isnewpara && !isemptyline)
ali@40	3020	{
ali@40	3021	/* This line is the start of a new paragraph. */
ali@68	3022	start_para_line=linecnt;
ali@40	3023	/* Capture its first line in case we want to report it later. */
ali@69	3024	g_free(parastart);
ali@69	3025	parastart=g_strdup(aline);
ali@56	3026	memset(&parities,0,sizeof(parities)); /* restart the quote count */
ali@68	3027	s=aline;
ali@70	3028	while (*s && !g_unichar_isalpha(g_utf8_get_char(s)) &&
ali@70	3029	!g_unichar_isdigit(g_utf8_get_char(s)))
ali@70	3030	s=g_utf8_next_char(s);
ali@70	3031	if (g_unichar_islower(g_utf8_get_char(s)))
ali@40	3032	{
ali@40	3033	/* and its first letter is lowercase */
ali@68	3034	if (pswit[ECHO_SWITCH])
ali@70	3035	g_print("\n%s\n",aline);
ali@68	3036	if (!pswit[OVERVIEW_SWITCH])
ali@70	3037	g_print(" Line %ld column %ld - "
ali@40	3038	"Paragraph starts with lower-case\n",
ali@70	3039	linecnt,g_utf8_pointer_to_offset(aline,s)+1);
ali@68	3040	else
ali@68	3041	cnt_punct++;
ali@40	3042	}
ali@69	3043	isnewpara=FALSE; /* Signal the end of new para processing. */
ali@40	3044	}
ali@68	3045	/* Check for an em-dash broken at line end. */
ali@70	3046	if (enddash && g_utf8_get_char(aline)=='-')
ali@40	3047	{
ali@68	3048	if (pswit[ECHO_SWITCH])
ali@70	3049	g_print("\n%s\n",aline);
ali@68	3050	if (!pswit[OVERVIEW_SWITCH])
ali@70	3051	g_print(" Line %ld column 1 - Broken em-dash?\n",linecnt);
ali@68	3052	else
ali@68	3053	cnt_punct++;
ali@40	3054	}
ali@69	3055	enddash=FALSE;
ali@70	3056	for (s=g_utf8_prev_char(aline+strlen(aline));
ali@70	3057	g_utf8_get_char(s)==' ' && s>aline;s=g_utf8_prev_char(s))
ali@40	3058	;
ali@70	3059	if (s>=aline && g_utf8_get_char(s)=='-')
ali@69	3060	enddash=TRUE;
ali@67	3061	check_for_control_characters(aline);
ali@185	3062	check_for_odd_characters(aline,warnings,isemptyline);
ali@68	3063	if (warnings->longline)
ali@45	3064	check_for_long_line(aline);
ali@68	3065	if (warnings->shortline)
ali@45	3066	check_for_short_line(aline,&last);
ali@68	3067	last.blen=last.len;
ali@70	3068	last.len=g_utf8_strlen(aline,-1);
ali@70	3069	last.start=g_utf8_get_char(aline);
ali@46	3070	check_for_starting_punctuation(aline);
ali@68	3071	if (warnings->dash)
ali@40	3072	{
ali@47	3073	check_for_spaced_emdash(aline);
ali@47	3074	check_for_spaced_dash(aline);
ali@40	3075	}
ali@48	3076	check_for_unmarked_paragraphs(aline);
ali@49	3077	check_for_jeebies(aline);
ali@50	3078	check_for_mta_from(aline);
ali@51	3079	check_for_orphan_character(aline);
ali@52	3080	check_for_pling_scanno(aline);
ali@53	3081	check_for_extra_period(aline,warnings);
ali@54	3082	check_for_following_punctuation(aline);
ali@55	3083	check_for_typos(aline,warnings);
ali@56	3084	check_for_misspaced_punctuation(aline,&parities,isemptyline);
ali@57	3085	check_for_double_punctuation(aline,warnings);
ali@58	3086	check_for_spaced_quotes(aline);
ali@59	3087	check_for_miscased_genative(aline);
ali@60	3088	check_end_of_line(aline,warnings);
ali@61	3089	check_for_unspaced_bracket(aline);
ali@68	3090	if (warnings->endquote)
ali@62	3091	check_for_unpunctuated_endquote(aline);
ali@63	3092	check_for_html_tag(aline);
ali@64	3093	check_for_html_entity(aline);
ali@68	3094	if (isemptyline)
ali@40	3095	{
ali@65	3096	check_for_mismatched_quotes(&counters,&pending);
ali@103	3097	counters_reset(&counters);
ali@40	3098	/* let the next iteration know that it's starting a new para */
ali@69	3099	isnewpara=TRUE;
ali@69	3100	if (prevline)
ali@69	3101	check_for_omitted_punctuation(prevline,&last,start_para_line);
ali@40	3102	}
ali@69	3103	g_free(prevline);
ali@69	3104	prevline=g_strdup(aline);
ali@0	3105	}
ali@103	3106	linecnt++;
ali@103	3107	check_for_mismatched_quotes(&counters,&pending);
ali@103	3108	print_pending(NULL,parastart,&pending);
ali@103	3109	reset_pending(&pending);
ali@69	3110	if (prevline)
ali@69	3111	{
ali@69	3112	g_free(prevline);
ali@69	3113	prevline=NULL;
ali@69	3114	}
ali@69	3115	g_free(parastart);
ali@69	3116	g_free(prevline);
ali@69	3117	g_free(etext);
ali@79	3118	if (!pswit[OVERVIEW_SWITCH] && !pswit[VERBOSE_SWITCH])
ali@69	3119	g_tree_foreach(qword,report_duplicate_queries,NULL);
ali@69	3120	g_tree_unref(qword);
ali@69	3121	g_tree_unref(qperiod);
ali@99	3122	counters_destroy(&counters);
ali@70	3123	g_set_print_handler(NULL);
ali@70	3124	print_as_windows_1252(NULL);
ali@71	3125	if (pswit[MARKUP_SWITCH])
ali@71	3126	loseentities(NULL);
ali@0	3127	}
ali@0	3128
ali@40	3129	/*
ali@40	3130	* flgets:
ali@40	3131	*
ali@69	3132	* Get one line from the input text, checking for
ali@40	3133	* the existence of exactly one CR/LF line-end per line.
ali@40	3134	*
ali@40	3135	* Returns: a pointer to the line.
ali@40	3136	*/
ali@69	3137	char flgets(char *etext,long lcnt)
ali@0	3138	{
ali@70	3139	gunichar c;
ali@69	3140	gboolean isCR=FALSE;
ali@69	3141	char theline=etext;
ali@70	3142	char *eos=theline;
ali@70	3143	gchar *s;
ali@70	3144	for (;;)
ali@40	3145	{
ali@70	3146	c=g_utf8_get_char(*etext);
ali@173	3147	if (!c)
ali@173	3148	{
ali@173	3149	if (*etext==theline)
ali@173	3150	return NULL;
ali@173	3151	else if (pswit[LINE_END_SWITCH])
ali@173	3152	{
ali@173	3153	if (pswit[ECHO_SWITCH])
ali@173	3154	{
ali@173	3155	s=g_strndup(theline,eos-theline);
ali@173	3156	g_print("\n%s\n",s);
ali@173	3157	g_free(s);
ali@173	3158	}
ali@173	3159	if (!pswit[OVERVIEW_SWITCH])
ali@173	3160	/* There may, or may not, have been a CR */
ali@173	3161	g_print(" Line %ld - No LF?\n",lcnt);
ali@173	3162	else
ali@173	3163	cnt_lineend++;
ali@173	3164	}
ali@173	3165	break;
ali@173	3166	}
ali@70	3167	etext=g_utf8_next_char(etext);
ali@40	3168	/* either way, it's end of line */
ali@69	3169	if (c=='\n')
ali@40	3170	{
ali@68	3171	if (isCR)
ali@68	3172	break;
ali@68	3173	else
ali@40	3174	{
ali@40	3175	/* Error - a LF without a preceding CR */
ali@68	3176	if (pswit[LINE_END_SWITCH])
ali@40	3177	{
ali@68	3178	if (pswit[ECHO_SWITCH])
ali@70	3179	{
ali@70	3180	s=g_strndup(theline,eos-theline);
ali@70	3181	g_print("\n%s\n",s);
ali@70	3182	g_free(s);
ali@70	3183	}
ali@68	3184	if (!pswit[OVERVIEW_SWITCH])
ali@70	3185	g_print(" Line %ld - No CR?\n",lcnt);
ali@68	3186	else
ali@68	3187	cnt_lineend++;
ali@40	3188	}
ali@68	3189	break;
ali@40	3190	}
ali@40	3191	}
ali@69	3192	if (c=='\r')
ali@40	3193	{
ali@68	3194	if (isCR)
ali@40	3195	{
ali@40	3196	/* Error - two successive CRs */
ali@68	3197	if (pswit[LINE_END_SWITCH])
ali@40	3198	{
ali@68	3199	if (pswit[ECHO_SWITCH])
ali@70	3200	{
ali@70	3201	s=g_strndup(theline,eos-theline);
ali@70	3202	g_print("\n%s\n",s);
ali@70	3203	g_free(s);
ali@70	3204	}
ali@68	3205	if (!pswit[OVERVIEW_SWITCH])
ali@70	3206	g_print(" Line %ld - Two successive CRs?\n",lcnt);
ali@68	3207	else
ali@68	3208	cnt_lineend++;
ali@40	3209	}
ali@40	3210	}
ali@69	3211	isCR=TRUE;
ali@40	3212	}
ali@68	3213	else
ali@40	3214	{
ali@68	3215	if (pswit[LINE_END_SWITCH] && isCR)
ali@40	3216	{
ali@68	3217	if (pswit[ECHO_SWITCH])
ali@70	3218	{
ali@70	3219	s=g_strndup(theline,eos-theline);
ali@70	3220	g_print("\n%s\n",s);
ali@70	3221	g_free(s);
ali@70	3222	}
ali@68	3223	if (!pswit[OVERVIEW_SWITCH])
ali@70	3224	g_print(" Line %ld column %ld - CR without LF?\n",
ali@70	3225	lcnt,g_utf8_pointer_to_offset(theline,eos)+1);
ali@68	3226	else
ali@68	3227	cnt_lineend++;
ali@70	3228	*eos=' ';
ali@40	3229	}
ali@69	3230	isCR=FALSE;
ali@70	3231	eos=g_utf8_next_char(eos);
ali@40	3232	}
ali@69	3233	}
ali@70	3234	*eos='\0';
ali@0	3235	if (pswit[MARKUP_SWITCH])
ali@68	3236	postprocess_for_HTML(theline);
ali@0	3237	if (pswit[DP_SWITCH])
ali@68	3238	postprocess_for_DP(theline);
ali@40	3239	return theline;
ali@0	3240	}
ali@0	3241
ali@40	3242	/*
ali@40	3243	* mixdigit:
ali@40	3244	*
ali@40	3245	* Takes a "word" as a parameter, and checks whether it
ali@40	3246	* contains a mixture of alpha and digits. Generally, this is an
ali@40	3247	* error, but may not be for cases like 4th or L5 12s. 3d.
ali@40	3248	*
ali@70	3249	* Returns: TRUE iff an is error found.
ali@40	3250	*/
ali@70	3251	gboolean mixdigit(const char *checkword)
ali@0	3252	{
ali@70	3253	gboolean wehaveadigit,wehavealetter,query;
ali@70	3254	const char s,nondigit;
ali@70	3255	wehaveadigit=wehavealetter=query=FALSE;
ali@70	3256	for (s=checkword;*s;s=g_utf8_next_char(s))
ali@70	3257	if (g_unichar_isalpha(g_utf8_get_char(s)))
ali@70	3258	wehavealetter=TRUE;
ali@70	3259	else if (g_unichar_isdigit(g_utf8_get_char(s)))
ali@70	3260	wehaveadigit=TRUE;
ali@40	3261	if (wehaveadigit && wehavealetter)
ali@40	3262	{
ali@40	3263	/* Now exclude common legit cases, like "21st" and "12l. 3s. 11d." */
ali@70	3264	query=TRUE;
ali@70	3265	for (nondigit=checkword;g_unichar_isdigit(g_utf8_get_char(nondigit));
ali@70	3266	nondigit=g_utf8_next_char(nondigit))
ali@68	3267	;
ali@68	3268	/* digits, ending in st, rd, nd, th of either case */
ali@70	3269	if (!g_ascii_strcasecmp(nondigit,"st") \|\|
ali@70	3270	!g_ascii_strcasecmp(nondigit,"rd") \|\|
ali@70	3271	!g_ascii_strcasecmp(nondigit,"nd") \|\|
ali@70	3272	!g_ascii_strcasecmp(nondigit,"th"))
ali@70	3273	query=FALSE;
ali@70	3274	if (!g_ascii_strcasecmp(nondigit,"sts") \|\|
ali@70	3275	!g_ascii_strcasecmp(nondigit,"rds") \|\|
ali@70	3276	!g_ascii_strcasecmp(nondigit,"nds") \|\|
ali@70	3277	!g_ascii_strcasecmp(nondigit,"ths"))
ali@70	3278	query=FALSE;
ali@70	3279	if (!g_ascii_strcasecmp(nondigit,"stly") \|\|
ali@70	3280	!g_ascii_strcasecmp(nondigit,"rdly") \|\|
ali@70	3281	!g_ascii_strcasecmp(nondigit,"ndly") \|\|
ali@70	3282	!g_ascii_strcasecmp(nondigit,"thly"))
ali@70	3283	query=FALSE;
ali@68	3284	/* digits, ending in l, L, s or d */
ali@70	3285	if (!g_ascii_strcasecmp(nondigit,"l") \|\| !strcmp(nondigit,"s") \|\|
ali@70	3286	!strcmp(nondigit,"d"))
ali@70	3287	query=FALSE;
ali@68	3288	/*
ali@40	3289	* L at the start of a number, representing Britsh pounds, like L500.
ali@70	3290	* This is cute. We know the current word is mixed digit. If the first
ali@68	3291	* letter is L, there must be at least one digit following. If both
ali@68	3292	* digits and letters follow, we have a genuine error, else we have a
ali@68	3293	* capital L followed by digits, and we accept that as a non-error.
ali@40	3294	*/
ali@70	3295	if (g_utf8_get_char(checkword)=='L' &&
ali@70	3296	!mixdigit(g_utf8_next_char(checkword)))
ali@70	3297	query=FALSE;
ali@40	3298	}
ali@40	3299	return query;
ali@0	3300	}
ali@0	3301
ali@40	3302	/*
ali@40	3303	* getaword:
ali@40	3304	*
ali@69	3305	* Extracts the first/next "word" from the line, and returns it.
ali@69	3306	* A word is defined as one English word unit--or at least that's the aim.
ali@69	3307	* "ptr" is advanced to the position in the line where we will start
ali@69	3308	* looking for the next word.
ali@40	3309	*
ali@69	3310	* Returns: A newly-allocated string.
ali@40	3311	*/
ali@69	3312	gchar getaword(const char *ptr)
ali@0	3313	{
ali@70	3314	const char s,t;
ali@69	3315	GString *word;
ali@70	3316	gunichar c,pc;
ali@69	3317	word=g_string_new(NULL);
ali@70	3318	for (;!g_unichar_isdigit(g_utf8_get_char(*ptr)) &&
ali@70	3319	!g_unichar_isalpha(g_utf8_get_char(*ptr)) &&
ali@70	3320	*ptr;ptr=g_utf8_next_char(*ptr))
ali@174	3321	{
ali@174	3322	/* Handle exceptions for footnote markers like [1] */
ali@174	3323	if (g_utf8_get_char(*ptr)=='[')
ali@174	3324	{
ali@174	3325	g_string_append_c(word,'[');
ali@174	3326	s=g_utf8_next_char(*ptr);
ali@174	3327	for (;g_unichar_isdigit(g_utf8_get_char(s));s=g_utf8_next_char(s))
ali@174	3328	g_string_append_unichar(word,g_utf8_get_char(s));
ali@174	3329	if (g_utf8_get_char(s)==']')
ali@174	3330	{
ali@174	3331	g_string_append_c(word,']');
ali@174	3332	*ptr=g_utf8_next_char(s);
ali@174	3333	return g_string_free(word,FALSE);
ali@174	3334	}
ali@174	3335	else
ali@174	3336	g_string_truncate(word,0);
ali@174	3337	}
ali@174	3338	}
ali@40	3339	/*
ali@40	3340	* Use a look-ahead to handle exceptions for numbers like 1,000 and 1.35.
ali@40	3341	* Especially yucky is the case of L1,000
ali@40	3342	* This section looks for a pattern of characters including a digit
ali@40	3343	* followed by a comma or period followed by one or more digits.
ali@40	3344	* If found, it returns this whole pattern as a word; otherwise we discard
ali@40	3345	* the results and resume our normal programming.
ali@40	3346	*/
ali@69	3347	s=*ptr;
ali@70	3348	for (;g_unichar_isdigit(g_utf8_get_char(s)) \|\|
ali@70	3349	g_unichar_isalpha(g_utf8_get_char(s)) \|\|
ali@70	3350	g_utf8_get_char(s)==',' \|\| g_utf8_get_char(s)=='.';s=g_utf8_next_char(s))
ali@70	3351	g_string_append_unichar(word,g_utf8_get_char(s));
ali@82	3352	if (word->len)
ali@40	3353	{
ali@82	3354	for (t=g_utf8_next_char(word->str);*t;t=g_utf8_next_char(t))
ali@40	3355	{
ali@82	3356	c=g_utf8_get_char(t);
ali@82	3357	pc=g_utf8_get_char(g_utf8_prev_char(t));
ali@82	3358	if ((c=='.' \|\| c==',') && g_unichar_isdigit(pc))
ali@82	3359	{
ali@82	3360	*ptr=s;
ali@82	3361	return g_string_free(word,FALSE);
ali@82	3362	}
ali@40	3363	}
ali@40	3364	}
ali@0	3365	/* we didn't find a punctuated number - do the regular getword thing */
ali@69	3366	g_string_truncate(word,0);
ali@99	3367	c=g_utf8_get_char(*ptr);
ali@99	3368	for (;g_unichar_isdigit(c) \|\| g_unichar_isalpha(c) \|\| CHAR_IS_APOSTROPHE(c);
ali@99	3369	ptr=g_utf8_next_char(ptr),c=g_utf8_get_char(*ptr))
ali@99	3370	g_string_append_unichar(word,c);
ali@69	3371	return g_string_free(word,FALSE);
ali@0	3372	}
ali@0	3373
ali@40	3374	/*
ali@40	3375	* isroman:
ali@40	3376	*
ali@40	3377	* Is this word a Roman Numeral?
ali@40	3378	*
ali@40	3379	* It doesn't actually validate that the number is a valid Roman Numeral--for
ali@40	3380	* example it will pass MXXXXXXXXXX as a valid Roman Numeral, but that's not
ali@40	3381	* what we're here to do. If it passes this, it LOOKS like a Roman numeral.
ali@40	3382	* Anyway, the actual Romans were pretty tolerant of bad arithmetic, or
ali@40	3383	* expressions thereof, except when it came to taxes. Allow any number of M,
ali@40	3384	* an optional D, an optional CM or CD, any number of optional Cs, an optional
ali@40	3385	* XL or an optional XC, an optional IX or IV, an optional V and any number
ali@40	3386	* of optional Is.
ali@40	3387	*/
ali@69	3388	gboolean isroman(const char *t)
ali@0	3389	{
ali@69	3390	const char *s;
ali@40	3391	if (!t \|\| !*t)
ali@69	3392	return FALSE;
ali@40	3393	s=t;
ali@70	3394	while (g_utf8_get_char(t)=='m' && *t)
ali@40	3395	t++;
ali@70	3396	if (g_utf8_get_char(t)=='d')
ali@40	3397	t++;
ali@70	3398	if (g_str_has_prefix(t,"cm"))
ali@40	3399	t+=2;
ali@70	3400	if (g_str_has_prefix(t,"cd"))
ali@40	3401	t+=2;
ali@70	3402	while (g_utf8_get_char(t)=='c' && *t)
ali@40	3403	t++;
ali@70	3404	if (g_str_has_prefix(t,"xl"))
ali@40	3405	t+=2;
ali@70	3406	if (g_str_has_prefix(t,"xc"))
ali@40	3407	t+=2;
ali@70	3408	if (g_utf8_get_char(t)=='l')
ali@40	3409	t++;
ali@70	3410	while (g_utf8_get_char(t)=='x' && *t)
ali@40	3411	t++;
ali@70	3412	if (g_str_has_prefix(t,"ix"))
ali@40	3413	t+=2;
ali@70	3414	if (g_str_has_prefix(t,"iv"))
ali@40	3415	t+=2;
ali@70	3416	if (g_utf8_get_char(t)=='v')
ali@40	3417	t++;
ali@70	3418	while (g_utf8_get_char(t)=='i' && *t)
ali@40	3419	t++;
ali@40	3420	return !*t;
ali@0	3421	}
ali@0	3422
ali@40	3423	/*
ali@40	3424	* postprocess_for_DP:
ali@40	3425	*
ali@40	3426	* Invoked with the -d switch from flgets().
ali@40	3427	* It simply "removes" from the line a hard-coded set of common
ali@40	3428	* DP-specific tags, so that the line passed to the main routine has
ali@40	3429	* been pre-cleaned of DP markup.
ali@40	3430	*/
ali@0	3431	void postprocess_for_DP(char *theline)
ali@0	3432	{
ali@40	3433	char s,t;
ali@0	3434	int i;
ali@0	3435	if (!*theline)
ali@68	3436	return;
ali@40	3437	for (i=0;*DPmarkup[i];i++)
ali@70	3438	while ((s=strstr(theline,DPmarkup[i])))
ali@40	3439	{
ali@68	3440	t=s+strlen(DPmarkup[i]);
ali@70	3441	memmove(s,t,strlen(t)+1);
ali@40	3442	}
ali@0	3443	}
ali@0	3444
ali@40	3445	/*
ali@40	3446	* postprocess_for_HTML:
ali@40	3447	*
ali@40	3448	* Invoked with the -m switch from flgets().
ali@40	3449	* It simply "removes" from the line a hard-coded set of common
ali@40	3450	* HTML tags and "replaces" a hard-coded set of common HTML
ali@40	3451	* entities, so that the line passed to the main routine has
ali@40	3452	* been pre-cleaned of HTML.
ali@40	3453	*/
ali@0	3454	void postprocess_for_HTML(char *theline)
ali@0	3455	{
ali@70	3456	while (losemarkup(theline))
ali@70	3457	;
ali@71	3458	loseentities(theline);
ali@0	3459	}
ali@0	3460
ali@0	3461	char losemarkup(char theline)
ali@0	3462	{
ali@40	3463	char s,t;
ali@0	3464	int i;
ali@70	3465	s=strchr(theline,'<');
ali@70	3466	t=s?strchr(s,'>'):NULL;
ali@40	3467	if (!s \|\| !t)
ali@40	3468	return NULL;
ali@40	3469	for (i=0;*markup[i];i++)
ali@70	3470	if (tagcomp(g_utf8_next_char(s),markup[i]))
ali@40	3471	{
ali@70	3472	t=g_utf8_next_char(t);
ali@70	3473	memmove(s,t,strlen(t)+1);
ali@70	3474	return s;
ali@68	3475	}
ali@40	3476	/* It's an unrecognized <xxx>. */
ali@40	3477	return NULL;
ali@0	3478	}
ali@0	3479
ali@71	3480	void loseentities(char *theline)
ali@0	3481	{
ali@0	3482	int i;
ali@71	3483	gsize nb;
ali@71	3484	char amp,scolon;
ali@71	3485	gchar s,t;
ali@71	3486	gunichar c;
ali@71	3487	GTree *entities=NULL;
ali@86	3488	static GIConv translit=(GIConv)-1,to_utf8=(GIConv)-1;
ali@71	3489	if (!theline)
ali@40	3490	{
ali@71	3491	if (entities)
ali@71	3492	g_tree_destroy(entities);
ali@71	3493	entities=NULL;
ali@86	3494	if (translit!=(GIConv)-1)
ali@71	3495	g_iconv_close(translit);
ali@71	3496	translit=(GIConv)-1;
ali@86	3497	if (to_utf8!=(GIConv)-1)
ali@71	3498	g_iconv_close(to_utf8);
ali@71	3499	to_utf8=(GIConv)-1;
ali@71	3500	return;
ali@71	3501	}
ali@71	3502	if (!*theline)
ali@71	3503	return;
ali@71	3504	if (!entities)
ali@71	3505	{
ali@71	3506	entities=g_tree_new((GCompareFunc)strcmp);
ali@71	3507	for(i=0;i<G_N_ELEMENTS(HTMLentities);i++)
ali@71	3508	g_tree_insert(entities,HTMLentities[i].name,
ali@71	3509	GUINT_TO_POINTER(HTMLentities[i].c));
ali@71	3510	}
ali@71	3511	if (translit==(GIConv)-1)
ali@71	3512	translit=g_iconv_open("ISO_8859-1//TRANSLIT","UTF-8");
ali@71	3513	if (to_utf8==(GIConv)-1)
ali@71	3514	to_utf8=g_iconv_open("UTF-8","ISO_8859-1");
ali@71	3515	while((amp=strchr(theline,'&')))
ali@71	3516	{
ali@71	3517	scolon=strchr(amp,';');
ali@71	3518	if (scolon)
ali@40	3519	{
ali@71	3520	if (amp[1]=='#')
ali@71	3521	{
ali@71	3522	if (amp+2+strspn(amp+2,"0123456789")==scolon)
ali@71	3523	c=strtol(amp+2,NULL,10);
ali@71	3524	else if (amp[2]=='x' &&
ali@71	3525	amp+3+strspn(amp+3,"0123456789abcdefABCDEF")==scolon)
ali@71	3526	c=strtol(amp+3,NULL,16);
ali@71	3527	}
ali@71	3528	else
ali@71	3529	{
ali@71	3530	s=g_strndup(amp+1,scolon-(amp+1));
ali@71	3531	c=GPOINTER_TO_UINT(g_tree_lookup(entities,s));
ali@71	3532	g_free(s);
ali@71	3533	}
ali@40	3534	}
ali@71	3535	else
ali@71	3536	c=0;
ali@71	3537	if (c)
ali@71	3538	{
ali@71	3539	theline=amp;
ali@71	3540	if (c<128 \|\| c>=192 && c<=255) /* An ISO-8859-1 character */
ali@71	3541	theline+=g_unichar_to_utf8(c,theline);
ali@71	3542	else
ali@71	3543	{
ali@71	3544	s=g_malloc(6);
ali@71	3545	nb=g_unichar_to_utf8(c,s);
ali@71	3546	t=g_convert_with_iconv(s,nb,translit,NULL,&nb,NULL);
ali@71	3547	g_free(s);
ali@71	3548	s=g_convert_with_iconv(t,nb,to_utf8,NULL,&nb,NULL);
ali@71	3549	g_free(t);
ali@71	3550	memcpy(theline,s,nb);
ali@71	3551	g_free(s);
ali@71	3552	theline+=nb;
ali@71	3553	}
ali@71	3554	memmove(theline,g_utf8_next_char(scolon),
ali@71	3555	strlen(g_utf8_next_char(scolon))+1);
ali@71	3556	}
ali@71	3557	else
ali@71	3558	theline=g_utf8_next_char(amp);
ali@40	3559	}
ali@0	3560	}
ali@0	3561
ali@70	3562	gboolean tagcomp(const char strin,const char basetag)
ali@0	3563	{
ali@70	3564	gboolean retval;
ali@70	3565	gchar s,t;
ali@70	3566	if (g_utf8_get_char(strin)=='/')
ali@70	3567	t=g_utf8_casefold(g_utf8_next_char(strin),-1); /* ignore a slash */
ali@70	3568	else
ali@70	3569	t=g_utf8_casefold(strin,-1);
ali@70	3570	s=g_utf8_casefold(basetag,-1);
ali@70	3571	retval=g_str_has_prefix(t,s);
ali@70	3572	g_free(s);
ali@70	3573	g_free(t);
ali@70	3574	return retval;
ali@0	3575	}
ali@0	3576
ali@69	3577	void proghelp(GOptionContext *context)
ali@0	3578	{
ali@69	3579	gchar *help;
ali@40	3580	fputs("Bookloupe version " PACKAGE_VERSION ".\n",stderr);
ali@40	3581	fputs("Copyright 2000-2005 Jim Tinsley <jtinsley@pobox.com>.\n",stderr);
ali@40	3582	fputs("Copyright 2012- J. Ali Harlow <ali@juiblex.co.uk>.\n",stderr);
ali@40	3583	fputs("Bookloupe comes wih ABSOLUTELY NO WARRANTY. "
ali@40	3584	"For details, read the file COPYING.\n",stderr);
ali@40	3585	fputs("This is Free Software; "
ali@40	3586	"you may redistribute it under certain conditions (GPL);\n",stderr);
ali@40	3587	fputs("read the file COPYING for details.\n\n",stderr);
ali@69	3588	help=g_option_context_get_help(context,TRUE,NULL);
ali@69	3589	fputs(help,stderr);
ali@69	3590	g_free(help);
ali@69	3591	fputs("Sample usage: bookloupe warpeace.txt\n\n",stderr);
ali@40	3592	fputs("Bookloupe queries anything it thinks shouldn't be in a PG text; "
ali@40	3593	"non-ASCII\n",stderr);
ali@40	3594	fputs("characters like accented letters, "
ali@40	3595	"lines longer than 75 or shorter than 55,\n",stderr);
ali@40	3596	fputs("unbalanced quotes or brackets, "
ali@40	3597	"a variety of badly formatted punctuation, \n",stderr);
ali@40	3598	fputs("HTML tags, some likely typos. "
ali@40	3599	"It is NOT a substitute for human judgement.\n",stderr);
ali@0	3600	fputs("\n",stderr);
ali@0	3601	}

author	ali <ali@juiblex.co.uk>
	Mon Oct 21 23:39:54 2013 +0100 (2013-10-21)
changeset 192	1aeda7fe17ca
parent 191	189183b37598
parent 185	a6d93c9932ac
child 193	7fdf168fb748
permissions	-rw-r--r--