1.1 --- a/bookloupe/bookloupe.c Sat Oct 05 21:37:31 2013 +0100
1.2 +++ b/bookloupe/bookloupe.c Sun Sep 29 09:18:05 2013 +0100
1.3 @@ -32,6 +32,9 @@
1.4 #include "pending.h"
1.5 #include "HTMLentities.h"
1.6
1.7 +gchar *charset; /* Or NULL for auto (ISO_8859-1/ASCII or UNICODE) */
1.8 +GIConv charset_validator=(GIConv)-1;
1.9 +
1.10 gchar *prevline;
1.11
1.12 /* Common typos. */
1.13 @@ -127,6 +130,7 @@
1.14 };
1.15
1.16 gboolean pswit[SWITNO]; /* program switches */
1.17 +gchar *opt_charset;
1.18
1.19 static GOptionEntry options[]={
1.20 { "dp", 'd', 0, G_OPTION_ARG_NONE, pswit+DP_SWITCH,
1.21 @@ -157,6 +161,8 @@
1.22 "Defaults for use on www upload", NULL },
1.23 { "verbose", 'v', 0, G_OPTION_ARG_NONE, pswit+VERBOSE_SWITCH,
1.24 "Verbose - list everything", NULL },
1.25 + { "charset", 0, 0, G_OPTION_ARG_STRING, &opt_charset,
1.26 + "Set of characters valid for this ebook", "NAME" },
1.27 { NULL }
1.28 };
1.29
1.30 @@ -200,6 +206,49 @@
1.31 UINT saved_cp;
1.32 #endif
1.33
1.34 +gboolean set_charset(const char *name,GError **err)
1.35 +{
1.36 + /* The various UNICODE encodings all share the same character set. */
1.37 + const char *unicode_aliases[]={ "UCS-2", "UCS-2BE", "UCS-2LE", "UCS-4",
1.38 + "UCS-4BE", "UCS-4LE", "UCS2", "UCS4", "UNICODE", "UNICODEBIG",
1.39 + "UNICODELITTLE", "UTF-7", "UTF-8", "UTF-16", "UTF-16BE", "UTF-16LE",
1.40 + "UTF-32", "UTF-32BE", "UTF-32LE", "UTF7", "UTF8", "UTF16", "UTF16BE",
1.41 + "UTF16LE", "UTF32", "UTF32BE", "UTF32LE" };
1.42 + int i;
1.43 + if (charset)
1.44 + g_free(charset);
1.45 + if (charset_validator!=(GIConv)-1)
1.46 + g_iconv_close(charset_validator);
1.47 + if (!name || !g_strcasecmp(name,"auto"))
1.48 + {
1.49 + charset=NULL;
1.50 + charset_validator=(GIConv)-1;
1.51 + return TRUE;
1.52 + }
1.53 + else
1.54 + charset=g_strdup(name);
1.55 + for(i=0;i<G_N_ELEMENTS(unicode_aliases);i++)
1.56 + if (!g_strcasecmp(charset,unicode_aliases[i]))
1.57 + {
1.58 + g_free(charset);
1.59 + charset=g_strdup("UTF-8");
1.60 + break;
1.61 + }
1.62 + if (!strcmp(charset,"UTF-8"))
1.63 + charset_validator=(GIConv)-1;
1.64 + else
1.65 + {
1.66 + charset_validator=g_iconv_open(charset,"UTF-8");
1.67 + if (charset_validator==(GIConv)-1)
1.68 + {
1.69 + g_set_error(err,G_CONVERT_ERROR,G_CONVERT_ERROR_NO_CONVERSION,
1.70 + "Unknown character set \"%s\"",charset);
1.71 + return FALSE;
1.72 + }
1.73 + }
1.74 + return TRUE;
1.75 +}
1.76 +
1.77 void parse_options(int *argc,char ***argv)
1.78 {
1.79 GError *err=NULL;
1.80 @@ -246,6 +295,13 @@
1.81 pswit[USERTYPO_SWITCH]=FALSE;
1.82 pswit[DP_SWITCH]=FALSE;
1.83 }
1.84 + if (opt_charset && !set_charset(opt_charset,&err))
1.85 + {
1.86 + g_printerr("%s\n",err->message);
1.87 + exit(1);
1.88 + }
1.89 + g_free(opt_charset);
1.90 + opt_charset=NULL;
1.91 if (*argc<2)
1.92 {
1.93 proghelp(context);
1.94 @@ -305,7 +361,11 @@
1.95 exit(1);
1.96 }
1.97 if (g_utf8_validate(contents,len,NULL))
1.98 + {
1.99 utf8=g_utf8_normalize(contents,len,G_NORMALIZE_DEFAULT_COMPOSE);
1.100 + if (!charset)
1.101 + (void)set_charset("UNICODE",NULL);
1.102 + }
1.103 else
1.104 utf8=g_convert(contents,len,"UTF-8","WINDOWS-1252",NULL,&nb,NULL);
1.105 g_free(contents);
1.106 @@ -428,6 +488,7 @@
1.107 g_free(running_from);
1.108 if (usertypo)
1.109 g_tree_unref(usertypo);
1.110 + set_charset(NULL,NULL);
1.111 return 0;
1.112 }
1.113
1.114 @@ -740,25 +801,32 @@
1.115 "Not reporting them.\n",
1.116 results->spacedash+results->emdash.non_PG_space);
1.117 }
1.118 - /* If more than a quarter of characters are hi-bit, bug out. */
1.119 - warnings.bin=1;
1.120 - if (results->binlen*4>results->totlen)
1.121 + if (charset)
1.122 + warnings.bin=0;
1.123 + else
1.124 {
1.125 - g_print(" --> This file does not appear to be ASCII. "
1.126 - "Terminating. Best of luck with it!\n");
1.127 - exit(1);
1.128 - }
1.129 - if (results->alphalen*4<results->totlen)
1.130 - {
1.131 - g_print(" --> This file does not appear to be text. "
1.132 - "Terminating. Best of luck with it!\n");
1.133 - exit(1);
1.134 - }
1.135 - if (results->binlen*100>results->totlen || results->binlen>100)
1.136 - {
1.137 - g_print(" --> There are a lot of foreign letters here. "
1.138 - "Not reporting them.\n");
1.139 - warnings.bin=0;
1.140 + /* Charset ISO_8859-1/ASCII checks for compatibility with gutcheck */
1.141 + warnings.bin=1;
1.142 + /* If more than a quarter of characters are hi-bit, bug out. */
1.143 + if (results->binlen*4>results->totlen)
1.144 + {
1.145 + g_print(" --> This file does not appear to be ASCII. "
1.146 + "Terminating. Best of luck with it!\n");
1.147 + exit(1);
1.148 + }
1.149 + if (results->alphalen*4<results->totlen)
1.150 + {
1.151 + g_print(" --> This file does not appear to be text. "
1.152 + "Terminating. Best of luck with it!\n");
1.153 + exit(1);
1.154 + }
1.155 + if (results->binlen*100>results->totlen || results->binlen>100)
1.156 + {
1.157 + g_print(" --> There are a lot of foreign letters here. "
1.158 + "Not reporting them.\n");
1.159 + if (!pswit[VERBOSE_SWITCH])
1.160 + warnings.bin=0;
1.161 + }
1.162 }
1.163 warnings.isDutch=FALSE;
1.164 if (results->Dutchcount>50)
1.165 @@ -786,7 +854,6 @@
1.166 g_print("\n");
1.167 if (pswit[VERBOSE_SWITCH])
1.168 {
1.169 - warnings.bin=1;
1.170 warnings.shortline=1;
1.171 warnings.dotcomma=1;
1.172 warnings.longline=1;
1.173 @@ -981,14 +1048,17 @@
1.174 gboolean isemptyline)
1.175 {
1.176 /* Don't repeat multiple warnings on one line. */
1.177 - gboolean eNon_A=FALSE,eTab=FALSE,eTilde=FALSE;
1.178 + gboolean eInvalidChar=FALSE,eTab=FALSE,eTilde=FALSE;
1.179 gboolean eCarat=FALSE,eFSlash=FALSE,eAst=FALSE;
1.180 const char *s;
1.181 gunichar c;
1.182 + gsize nb;
1.183 + gchar *t;
1.184 for (s=aline;*s;s=g_utf8_next_char(s))
1.185 {
1.186 c=g_utf8_get_char(s);
1.187 - if (!eNon_A && (c<CHAR_SPACE && c!='\t' && c!='\n' || c>127))
1.188 + if (warnings->bin && !eInvalidChar &&
1.189 + (c<CHAR_SPACE && c!='\t' && c!='\n' || c>127))
1.190 {
1.191 if (pswit[ECHO_SWITCH])
1.192 g_print("\n%s\n",aline);
1.193 @@ -1003,7 +1073,57 @@
1.194 linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
1.195 else
1.196 cnt_bin++;
1.197 - eNon_A=TRUE;
1.198 + eInvalidChar=TRUE;
1.199 + }
1.200 + if (!eInvalidChar && charset)
1.201 + {
1.202 + if (charset_validator==(GIConv)-1)
1.203 + {
1.204 + if (!g_unichar_isdefined(c))
1.205 + {
1.206 + if (pswit[ECHO_SWITCH])
1.207 + g_print("\n%s\n",aline);
1.208 + if (!pswit[OVERVIEW_SWITCH])
1.209 + g_print(" Line %ld column %ld - Unassigned UNICODE "
1.210 + "code point U+%04" G_GINT32_MODIFIER "X\n",
1.211 + linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
1.212 + else
1.213 + cnt_bin++;
1.214 + eInvalidChar=TRUE;
1.215 + }
1.216 + else if (c>=0xE000 && c<=0xF8FF || c>=0xF0000 && c<=0xFFFFD ||
1.217 + c>=100000 && c<=0x10FFFD)
1.218 + {
1.219 + if (pswit[ECHO_SWITCH])
1.220 + g_print("\n%s\n",aline);
1.221 + if (!pswit[OVERVIEW_SWITCH])
1.222 + g_print(" Line %ld column %ld - Private Use "
1.223 + "character U+%04" G_GINT32_MODIFIER "X\n",
1.224 + linecnt,g_utf8_pointer_to_offset(aline,s)+1,c);
1.225 + else
1.226 + cnt_bin++;
1.227 + eInvalidChar=TRUE;
1.228 + }
1.229 + }
1.230 + else
1.231 + {
1.232 + t=g_convert_with_iconv(s,g_utf8_next_char(s)-s,
1.233 + charset_validator,NULL,&nb,NULL);
1.234 + if (t)
1.235 + g_free(t);
1.236 + else
1.237 + {
1.238 + if (pswit[ECHO_SWITCH])
1.239 + g_print("\n%s\n",aline);
1.240 + if (!pswit[OVERVIEW_SWITCH])
1.241 + g_print(" Line %ld column %ld - Non-%s "
1.242 + "character %u\n",linecnt,
1.243 + g_utf8_pointer_to_offset(aline,s)+1,charset,c);
1.244 + else
1.245 + cnt_bin++;
1.246 + eInvalidChar=TRUE;
1.247 + }
1.248 + }
1.249 }
1.250 if (!eTab && c==CHAR_TAB)
1.251 {
1.252 @@ -2691,8 +2811,7 @@
1.253 if (s>=aline && g_utf8_get_char(s)=='-')
1.254 enddash=TRUE;
1.255 check_for_control_characters(aline);
1.256 - if (warnings->bin)
1.257 - check_for_odd_characters(aline,warnings,isemptyline);
1.258 + check_for_odd_characters(aline,warnings,isemptyline);
1.259 if (warnings->longline)
1.260 check_for_long_line(aline);
1.261 if (warnings->shortline)
2.1 --- a/test/bookloupe/Makefile.am Sat Oct 05 21:37:31 2013 +0100
2.2 +++ b/test/bookloupe/Makefile.am Sun Sep 29 09:18:05 2013 +0100
2.3 @@ -1,6 +1,6 @@
2.4 TESTS_ENVIRONMENT=BOOKLOUPE=../../bookloupe/bookloupe ../harness/loupe-test
2.5 TESTS=non-ascii.tst long-line.tst curved-single-quotes.tst curved-quotes.tst \
2.6 runfox-quotes.tst curved-genitives.tst multi-line-illustration.tst \
2.7 - emdash.tst
2.8 + emdash.tst charset-cp1252.tst charset-latin1.tst
2.9
2.10 dist_pkgdata_DATA=$(TESTS)
3.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
3.2 +++ b/test/bookloupe/charset-cp1252.tst Sun Sep 29 09:18:05 2013 +0100
3.3 @@ -0,0 +1,16 @@
3.4 +**************** OPTIONS ****************
3.5 +--charset=WINDOWS-1252
3.6 +**************** ENCODING ****************
3.7 +WINDOWS-1252
3.8 +**************** INPUT ****************
3.9 +Unless binary mode is engaged, gutcheck will warn about a number of
3.10 +characters defined in Windows-1252. Bookloupe provides support for
3.11 +disabling such checks without concern as to the file size and how
3.12 +many characters with the eighth bit set it may contain by allowing a
3.13 +character set to be declared. With the character set declared as
3.14 +WINDOWS-1252, all characters defined in Windows-1252 shoud be acceptable
3.15 +and no warnings should be issued.
3.16 +
3.17 +We test for this by including just one such character—the em dash.
3.18 +
3.19 +**************** EXPECTED ****************
4.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
4.2 +++ b/test/bookloupe/charset-latin1.tst Sun Sep 29 09:18:05 2013 +0100
4.3 @@ -0,0 +1,58 @@
4.4 +**************** OPTIONS ****************
4.5 +--charset=ISO-8859-1
4.6 +**************** ENCODING ****************
4.7 +WINDOWS-1252
4.8 +**************** INPUT ****************
4.9 +Where the character set declared is narrower than the character set
4.10 +implied by the encoding as in this case (Windows-1252 is a superset
4.11 +of the first latin alphabet defined in ECMA 94), then bookloupe should
4.12 +warn about characters that are not in the declared character set but
4.13 +should still recognise them and otherwise handle them as it would
4.14 +normally do. We use the curved apostrophe as a test for this since
4.15 +if bookloupe didn't recognise it then it would query the orphaned
4.16 +letters from the genitives and abbreviations.
4.17 +
4.18 +John Hendricks was bear-leading at the time. He had originally studied
4.19 +for Holy Orders, but had abandoned the Church later for private reasons
4.20 +connected with his faith, and had taken to teaching and tutoring
4.21 +instead. He was an honest, upstanding fellow of five-and-thirty,
4.22 +incorruptible, intelligent in a simple, straightforward way. He played
4.23 +games with his head, more than most Englishmen do, but he went through
4.24 +life without much calculation. He had qualities that made boys like
4.25 +and respect him; he won their confidence. Poor, proud, ambitious,
4.26 +he realised that fate offered him a chance when the Secretary of
4.27 +State for Scotland asked him if he would give up his other pupils
4.28 +for a year and take his son, Lord Ernie, round the world upon an
4.29 +educational trip that might make a man of him. For Lord Ernie was the
4.30 +only son, and the Marquess’s influence was naturally great. To have
4.31 +deposited a regenerated Lord Ernie at the castle gates might have
4.32 +guaranteed Hendricks’ future. After leaving Eton prematurely the lad
4.33 +had come under Hendricks’ charge for a time, and with such excellent
4.34 +results--‘I’d simply swear by that chap, you know,’ the boy used
4.35 +to say--that his father, considerably impressed, and rather as a
4.36 +last resort, had made this proposition. And Hendricks, without much
4.37 +calculation, had accepted it. He liked ‘Bindy’ for himself. It was
4.38 +in his heart to ‘make a man of him,’ if possible. They had now been
4.39 +round the world together and had come up from Brindisi to the Italian
4.40 +Lakes, and so into Switzerland. It was middle October. With a week or
4.41 +two to spare they were making leisurely for the ancestral halls in
4.42 +Aberdeenshire.
4.43 +**************** EXPECTED ****************
4.44 +
4.45 +only son, and the Marquess’s influence was naturally great. To have
4.46 + Line 22 column 27 - Non-ISO-8859-1 character 8217
4.47 +
4.48 +guaranteed Hendricks’ future. After leaving Eton prematurely the lad
4.49 + Line 24 column 21 - Non-ISO-8859-1 character 8217
4.50 +
4.51 +had come under Hendricks’ charge for a time, and with such excellent
4.52 + Line 25 column 25 - Non-ISO-8859-1 character 8217
4.53 +
4.54 +results--‘I’d simply swear by that chap, you know,’ the boy used
4.55 + Line 26 column 10 - Non-ISO-8859-1 character 8216
4.56 +
4.57 +calculation, had accepted it. He liked ‘Bindy’ for himself. It was
4.58 + Line 29 column 40 - Non-ISO-8859-1 character 8216
4.59 +
4.60 +in his heart to ‘make a man of him,’ if possible. They had now been
4.61 + Line 30 column 17 - Non-ISO-8859-1 character 8216