- Dillo-dev - lists.mailman3.com

dillo CSS and charset

by jcid＠dillo.org

On Thu, Jan 22, 2009 at 12:24:41PM -0300, Jorge Arellano Cid wrote: > On Thu, Jan 22, 2009 at 11:06:01PM +0900, furaisanjin wrote: > > 2009/1/22 furaisanjin <furaisanjin(a)gmail.com>: > > > By the way I notice bookmark (ctrl-b) doesn't work. Does anybody has > > > this problem? > > > > bookmark problem is solved by this patch. > > > > --- a/src/cache.c Thu Jan 22 08:31:44 2009 -0300 > > +++ b/src/cache.c Thu Jan 22 23:03:33 2009 +0900 > > @@ -512,8 +512,10 @@ > > if (entry->CharsetDecoder) > > a_Decode_free(entry->CharsetDecoder); > > a_Misc_parse_content_type(ctype, NULL, NULL, &charset); > > - entry->CharsetDecoder = a_Decode_charset_init(charset); > > - dFree(charset); > > + if (charset) { > > + entry->CharsetDecoder = a_Decode_charset_init(charset); > > + dFree(charset); > > + } > > curr = Cache_current_content_type(entry); > > > > /* Invalidate UTF8Data */ > > Have you added charset to the content-type given by the bookmarks dpi? > or have you added a META or both? > > I would like to know the problem in detail, because it looks like > adding an extra parameter to a_Cache_set_content_type() would make > its API, and code cleaner. e.g. ({CA_Header|CA_Meta} as last parameter). > > If the problem can be solved with that API change, great, hence > the above question. Does something like the attached patch solve the problem? -- Cheers Jorge.-

Jan. 22, 2009

dillo CSS and charset

by corvid＠lavabit.com

Johannes wrote: > On Thu, Jan 22, 2009 at 05:57:42PM +0000, corvid wrote: > > Johannes wrote: > > > BTW. now you can set vw_fontname in dillorc again. This font will > > > then be used when CSS specifies "sans" as font, e.g: > > > vw_fontname="VL Gothic" > > > > If I put a serif font in vw_fontname and then use > > sans in style.css for some things, it'll use > > the serif font. > > True. We need to make new config options > > sans-serif_fontname > serif_fontname > cursive_fontname > fantasy_fontname > monospace_fontname > > maybe even dropping the "_fontname" suffix. > > But of course that would no longer be compatible with older > versions of dillorc. I would add the new options and remove the old options. CSS is big and certainly worth some incompatibility. Related to this, I've been wondering whether we should have a policy about the dillorc on the website -- whether it should follow tip, or follow the most recent release, or always add new options but never delete old options...

Jan. 22, 2009

dillo CSS and charset

by Johannes.Hofmann＠gmx.de

On Thu, Jan 22, 2009 at 05:57:42PM +0000, corvid wrote: > Johannes wrote: > > BTW. now you can set vw_fontname in dillorc again. This font will > > then be used when CSS specifies "sans" as font, e.g: > > vw_fontname="VL Gothic" > > If I put a serif font in vw_fontname and then use > sans in style.css for some things, it'll use > the serif font. True. We need to make new config options sans-serif_fontname serif_fontname cursive_fontname fantasy_fontname monospace_fontname maybe even dropping the "_fontname" suffix. But of course that would no longer be compatible with older versions of dillorc. Opinions? Cheers, Johannes

Jan. 22, 2009

dillo CSS and charset

by corvid＠lavabit.com

Johannes wrote: > BTW. now you can set vw_fontname in dillorc again. This font will > then be used when CSS specifies "sans" as font, e.g: > vw_fontname="VL Gothic" If I put a serif font in vw_fontname and then use sans in style.css for some things, it'll use the serif font.

Jan. 22, 2009

dillo CSS and charset

by jcid＠dillo.org

On Thu, Jan 22, 2009 at 11:06:01PM +0900, furaisanjin wrote: > 2009/1/22 furaisanjin <furaisanjin(a)gmail.com>: > > By the way I notice bookmark (ctrl-b) doesn't work. Does anybody has > > this problem? > > bookmark problem is solved by this patch. > > --- a/src/cache.c Thu Jan 22 08:31:44 2009 -0300 > +++ b/src/cache.c Thu Jan 22 23:03:33 2009 +0900 > @@ -512,8 +512,10 @@ > if (entry->CharsetDecoder) > a_Decode_free(entry->CharsetDecoder); > a_Misc_parse_content_type(ctype, NULL, NULL, &charset); > - entry->CharsetDecoder = a_Decode_charset_init(charset); > - dFree(charset); > + if (charset) { > + entry->CharsetDecoder = a_Decode_charset_init(charset); > + dFree(charset); > + } > curr = Cache_current_content_type(entry); > > /* Invalidate UTF8Data */ Have you added charset to the content-type given by the bookmarks dpi? or have you added a META or both? I would like to know the problem in detail, because it looks like adding an extra parameter to a_Cache_set_content_type() would make its API, and code cleaner. e.g. ({CA_Header|CA_Meta} as last parameter). If the problem can be solved with that API change, great, hence the above question. BTW AFAIS, with the above patch if CharsetDecoder was already set, and a_Misc_parse_content_type() fails, CharsetDecoder is left in an unknown state. -- Cheers Jorge.-

Jan. 22, 2009

dillo CSS and charset

by furaisanjin＠gmail.com

2009/1/22 furaisanjin <furaisanjin(a)gmail.com>: > By the way I notice bookmark (ctrl-b) doesn't work. Does anybody has > this problem? bookmark problem is solved by this patch. --- a/src/cache.c Thu Jan 22 08:31:44 2009 -0300 +++ b/src/cache.c Thu Jan 22 23:03:33 2009 +0900 @@ -512,8 +512,10 @@ if (entry->CharsetDecoder) a_Decode_free(entry->CharsetDecoder); a_Misc_parse_content_type(ctype, NULL, NULL, &charset); - entry->CharsetDecoder = a_Decode_charset_init(charset); - dFree(charset); + if (charset) { + entry->CharsetDecoder = a_Decode_charset_init(charset); + dFree(charset); + } curr = Cache_current_content_type(entry); /* Invalidate UTF8Data */ Regards, furaisanjin

Jan. 22, 2009

switch to DILLO_HTML_PARSE_MODE_PRE depending on style

by Johannes.Hofmann＠gmx.de

Hi, I'm experimenting with attached patch to make switching to DILLO_HTML_PARSE_MODE_PRE from CSS possible. Some online source browsers like cgit (http://hjemli.net/git/cgit/) use this to display sources or diffs. Please comment. Cheers, Johannes

Jan. 22, 2009

dillo CSS and charset

by Johannes.Hofmann＠gmx.de

On Thu, Jan 22, 2009 at 08:34:32AM -0300, Jorge Arellano Cid wrote: > On Thu, Jan 22, 2009 at 10:22:39AM +0100, Hofmann Johannes wrote: > > Hi, > > > > On Wed, Jan 21, 2009 at 07:06:06PM -0300, Jorge Arellano Cid wrote: > > > On Wed, Jan 21, 2009 at 08:29:34PM +0900, furaisanjin wrote: > > > > Hi. > > > > > > > > This is work around. > > > > > > This so called "work around" works very well. > > > > > > While working on this, I used the testcases in the following table: > > > > > > > > > load Reload Back > > > ---------------------------- > > > H__ OK OK OK H = has charset in HTTP header > > > HM_ OK OK OK M = has charset in META > > > H_C OK OK OK C = has remote CSS stylesheet > > > HMC OK OK OK _ = disabled > > > ___ OK OK OK > > > _M_ OK OK OK > > > __C OK OK OK > > > _MC OK OK OK > > > ---------------------------- > > > > > > Both patches pass all of them! > > > > > > I committed a similar approach that has a single code path for > > > calling a_Decode_charset_init(). > > > > > > Please test and comment. > > > > Starting with revision d29cdb5b842e I get repeatable crashes when > > clicking on the "Home" link in about:splash. > > Please try the current tip. I noticed one more issue, still present in tip. On http://amazon.de dillo crashes with: Program terminated with signal 11, Segmentation fault. b#0 0x08069f2e in Html_write_raw (html=0x2866f500, buf=0x2874303d <Address 0x2874303d out of bounds>, bufsize=33743, Eof=0) at html.cc:3398 3398 if (*start == '/' || /* </x> */ (gdb) bt #0 0x08069f2e in Html_write_raw (html=0x2866f500, buf=0x2874303d <Address 0x2874303d out of bounds>, bufsize=33743, Eof=0) at html.cc:3398 #1 0x0806a53a in DilloHtml::write (this=0x2866f500, Buf=0x28743000 <Address 0x28743000 out of bounds>, BufSize=675537188, Eof=0) at html.cc:577 #2 0x0806a58e in Html_callback (Op=0, Client=0x2862ea40) at html.cc:3564 #3 0x0805a706 in Cache_process_queue (entry=0x28653700) at cache.c:1143 #4 0x08058288 in a_Chain_fcb (Op=675537188, Info=0x1, Data1=0x286b8210, Data2=0x80f573c) at chain.c:112 #5 0x0807899c in a_Dpi_ccc (Op=2, Branch=2, Dir=1, Info=0x2862e400, Data1=0x286b8c20, Data2=0x0) at dpi.c:219 #6 0x08058288 in a_Chain_fcb (Op=675537188, Info=0x1, Data1=0x286b8c20, Data2=0x0) at chain.c:112 #7 0x08079311 in a_IO_ccc (Op=2, Branch=2, Dir=1, Info=0x2862ee20, Data1=0x2862e4a0, Data2=0x0) at IO.c:414 #8 0x0807949e in IO_callback (io=0x2862e4a0) at IO.c:201 #9 0x0807956f in IO_fd_read_cb (fd=4, data=0x4) at IO.c:287 #10 0x080c9fd5 in fltk::wait () #11 0x080ca185 in fltk::run () #12 0x0804f03a in main (argc=2, argv=0xbfbff850) at dillo.cc:334 Cheers, Johannes

Jan. 22, 2009

dillo CSS and charset

by furaisanjin＠gmail.com

Hi, 2009/1/22 Hofmann Johannes <Johannes.Hofmann(a)gmx.de>: > On Thu, Jan 22, 2009 at 09:18:00PM +0900, furaisanjin wrote: > However looking at japanese pages, it seems that word wrapping is > still an issue, see: > http://lists.auriga.wearlab.de/pipermail/dillo-dev/2008-November/005427.html > > Do you have an idea about a possible solution? No, I don't think there is good solution but I added these code in html.cc. --- a/src/html.cc Thu Jan 22 08:31:44 2009 -0300 +++ b/src/html.cc Thu Jan 22 21:33:56 2009 +0900 @@ -58,6 +58,8 @@ #define TAB_SIZE 8 +#define UTF8_NONE_1WORD(x) ((x)&0x80) + /*----------------------------------------------------------------------------- * Name spaces *---------------------------------------------------------------------------*/ @@ -111,6 +113,7 @@ DilloImage *image); static void Html_callback(int Op, CacheClient_t *Client); static void Html_tag_cleanup_at_close(DilloHtml *html, int TagIdx); +static int Html_utf8_size (char c); /*----------------------------------------------------------------------------- * Local Data @@ -1117,6 +1120,7 @@ { int i, j, start; char *Pword, ch; + int utf8size; DilloHtmlParseMode parse_mode = S_TOP(html)->parse_mode; if (parse_mode == DILLO_HTML_PARSE_MODE_STASH || @@ -1159,7 +1163,20 @@ } else { if (!memchr(word,'&', size)) { /* No entities */ - DW2TB(html->dw)->addText(word, html->styleEngine->wordStyle ()); + for (start = 0; start < (int) strlen (word); ) { + if (UTF8_NONE_1WORD(word[start])) { + utf8size = Html_utf8_size (word[start]); + DW2TB(html->dw)->addText(dStrndup(&word[start],utf8size), html->styleEngine->wordStyle()); + start += utf8size; + } else { + for (i=0; i < (int) strlen (word)-start; i++) { + if (UTF8_NONE_1WORD(word[start+i])) + break; + } + DW2TB(html->dw)->addText(dStrndup(&word[start],i), html->styleEngine->wordStyle()); + start += i; + } + } } else { /* Collapse white-space entities inside the word (except  ) */ Pword = a_Html_parse_entities(html, word, size); @@ -1167,7 +1184,20 @@ if (strchr("\t\f\n\r", Pword[i])) for (j = i; (Pword[j] = Pword[j+1]); ++j) ; - DW2TB(html->dw)->addText(Pword, html->styleEngine->wordStyle ()); + for (start = 0; start < (int) strlen (Pword); ) { + if (UTF8_NONE_1WORD(Pword[start])) { + utf8size = Html_utf8_size (Pword[start]); + DW2TB(html->dw)->addText(dStrndup(&Pword[start],utf8size), html->styleEngine->wordStyle()); + start += utf8size; + } else { + for (i=0; i < (int) strlen (Pword)-start; i++) { + if (UTF8_NONE_1WORD(Pword[start+i])) + break; + } + DW2TB(html->dw)->addText(dStrndup(&Pword[start],i), html->styleEngine->wordStyle()); + start += i; + } + } dFree(Pword); } } @@ -3700,4 +3730,20 @@ return token_start; } +static int Html_utf8_size (char c) +{ + if ((c&0xfc) == 0xfc) + return 6; + if ((c&0xf8) == 0xf8) + return 5; + if ((c&0xf0) == 0xf0) + return 4; + if ((c&0xe0) == 0xe0) + return 3; + if ((c&0xc0) == 0xc0) + return 2; + return 1; +} + Then dillo appearance is something I mentioned here. http://lists.auriga.wearlab.de/pipermail/dillo-dev/2008-November/005430.html By the way I notice bookmark (ctrl-b) doesn't work. Does anybody has this problem? Regards, furaisanjin

Jan. 22, 2009

dillo CSS and charset

by Johannes.Hofmann＠gmx.de

On Thu, Jan 22, 2009 at 09:18:00PM +0900, furaisanjin wrote: > Hi. > > And charset conversion works fine. BTW. now you can set vw_fontname in dillorc again. This font will then be used when CSS specifies "sans" as font, e.g: vw_fontname="VL Gothic" However looking at japanese pages, it seems that word wrapping is still an issue, see: http://lists.auriga.wearlab.de/pipermail/dillo-dev/2008-November/005427.html Do you have an idea about a possible solution? Cheers, Johannes

Jan. 22, 2009