Péter Szabó

Show:

What's happening?

  • tool.pdf.Compress increases PDF image size by 8%

    The latest tool.pdf.Compress increases the file size of the PDF http://www.math.bme.hu/~pts/300dpi_all.type1c.pdf by 8%. The most probable reason is that the input PDF contains lots of images, most of them being very tightly compressed with PNGOUT. When tool.pdf.Compress processes the PDF file, it recompresses the images with an inferior algorithm, not paying attention to the fact that the...

    2009-04-06 01:56:28 UTC in Multivalent

  • tool.pdf.Compress reads images with /Predictor incorrectly

    If the original image has a /DecodeParams dictionary with a /Predictor, tool.pdf.Compress ignores this dictionary, thus rendering the image completely useless. $ xpdf pts2p.pdf $ java -cp Multivalent.jar tool.pdf.Compress pts2p.pdf $ xpdf pts2p-o.pdf.

    2009-03-31 22:51:04 UTC in Multivalent

  • load builtin *joerc as a fallback if no *fancyjoerc (+patch)

    This patch makes joe load builtin *joerc as a fallback if builtin *fancyjoerc is not present. Index: main.c =================================================================== --- main.c (revision 38) +++ main.c (working copy) @@ -257,6 +257,14 @@ s = vsncpy(sv(s), sv(run)); s = vsncpy(sv(s), sc("rc")); c = procrc(cap, s); + if (c != 0 && c != 1) { + /* If built-in...

    2008-11-15 13:17:04 UTC in JOE - Joe's own editor

  • ustat() to show the byte length of UTF-8 char (with patch)

    This patch extends ustat() by showing the number of bytes the current character occupies. This information is useful for debugging UTF-8 text. Index: bw.c =================================================================== --- bw.c (revision 28) +++ bw.c (working copy) @@ -1273,8 +1273,15 @@ static unsigned char buf[160]; unsigned char bf1[100]; unsigned char bf2[100]; - int c...

    2008-11-15 13:00:27 UTC in JOE - Joe's own editor

  • speed up brch() in UTF-8 buffers

    This patch makes brch on UTF-8 buffers a little bit faster. Please apply after my earlier patch https://sourceforge.net/tracker/index.php?func=detail&aid=2290129&group_id=23475&atid=378598 Index: b.c =================================================================== --- b.c (revision 44) +++ b.c (working copy) @@ -2857,14 +2857,13 @@ int brch(P *p) { - if (p->b->o.charmap->type...

    2008-11-15 12:54:11 UTC in JOE - Joe's own editor

  • restore UTF-8 mode when leaving hex mode (with patch)

    This patch makes joe restore UTF-8 mode after entering hex mode in a UTF-8 buffer, and then leaving hex mode. A more general fix would restore the proper old o->charmap. Index: rc.c =================================================================== --- rc.c (revision 25) +++ rc.c (working copy) @@ -243,6 +243,7 @@ /* Hex not allowed with UTF-8 */ if (o->hex && o->charmap->type) {.

    2008-11-15 12:43:30 UTC in JOE - Joe's own editor

  • remove some rarely used math functions from the uClibc build

    This patch removes some rare math functions from a Linux uClibc build. The system library uClibc doesn't have those math functions, and we'd get a link error if we used them. A better fix would detect the presence of these math functions is ./configure. Index: umath.c =================================================================== --- umath.c (revision 5) +++ umath.c (working copy...

    2008-11-15 11:50:55 UTC in JOE - Joe's own editor

  • improved documentation of utf8_decode() (with patch)

    This patch improves documentation of utf8_decode(). Index: utf8.c =================================================================== --- utf8.c (revision 25) +++ utf8.c (working copy) @@ -86,9 +86,9 @@ /* UTF-8 Decoder * * Returns 0 - 7FFFFFFF: decoded character - * -1: character accepted, nothing decoded yet. - * -2: incomplete sequence - *.

    2008-11-15 11:42:25 UTC in JOE - Joe's own editor

  • documentation for pfwrd and pbkwd is wrong

    Here is a patch which fixes bad documentation comments of pfwrd and pbkwd. In addition to that, on http://joe-editor.sourceforge.net/hacking.html please change this: pfwrd(p,int n); Move forward n characters pbkwd(p,int n); Move backward n characters to this: pfwrd(p,long n); Move forward n bytes pbkwd(p,long n); Move backward n bytes Index: b.c...

    2008-11-15 11:20:20 UTC in JOE - Joe's own editor

  • Comment: Speed up b_goto_eol in UTF-8 and crlf buffers (with patch)

    Sorry, the previous one didn't update p->col properly. Here is the correct solution. This is a speed optimization which makes p_goto_eol faster on UTF-8 and crlf buffers. Our general observation is that we don't need special casing in case of UTF-8, because it keeps '\n', '\t' and '\n' intact. The only special casing we need in case of crlf is to stop at \r. Index: b.c...

    2008-11-15 11:08:23 UTC in JOE - Joe's own editor

About Me


Send me a message