Extract text of a pdf correctly
See "[poppler] text extraction does not work" in the mailing list for more info
diff --git a/poppler/GfxFont.cc b/poppler/GfxFont.cc
index 2d7180a..0a165a6 100644
--- a/poppler/GfxFont.cc
+++ b/poppler/GfxFont.cc
@@ -22,6 +22,7 @@
// Copyright (C) 2008, 2010 Hib Eris <hib@hiberis.nl>
// Copyright (C) 2009 Peter Kerzum <kerzum@yandex-team.ru>
// Copyright (C) 2009, 2010 David Benjamin <davidben@mit.edu>
+// Copyright (C) 2011 Axel Strübing <axel.struebing@freenet.de>
//
// To see a description of the changes please see the Changelog file that
// came with your tarball or type make ChangeLog if you are building from git
@@ -176,6 +177,7 @@
weight = WeightNotDefined;
refCnt = 1;
dfp = NULL;
+ hasToUnicode = gFalse;
}
GfxFont::~GfxFont() {
@@ -436,6 +438,7 @@
} else {
ctu = CharCodeToUnicode::parseCMap(buf, nBits);
}
+ hasToUnicode = gTrue;
delete buf;
return ctu;
}
@@ -1697,7 +1700,16 @@
*code = (CharCode)(cid = cMap->getCID(s, len, &n));
if (ctu) {
- *uLen = ctu->mapToUnicode(cid, u);
+ if (hasToUnicode) {
+ int i = 0, c = 0;
+ while (i < n) {
+ c = (c << 8 ) + (s[i] & 0xff);
+ ++i;
+ }
+ *uLen = ctu->mapToUnicode(c, u);
+ } else {
+ *uLen = ctu->mapToUnicode(cid, u);
+ }
} else {
*uLen = 0;
}
diff --git a/poppler/GfxFont.h b/poppler/GfxFont.h
index 55f8692..09af062 100644
--- a/poppler/GfxFont.h
+++ b/poppler/GfxFont.h
@@ -19,6 +19,7 @@
// Copyright (C) 2007 Julien Rebetez <julienr@svn.gnome.org>
// Copyright (C) 2007 Jeff Muizelaar <jeff@infidigm.net>
// Copyright (C) 2007 Koji Otani <sho@bbr.jp>
+// Copyright (C) 2011 Axel Strübing <axel.struebing@freenet.de>
//
// To see a description of the changes please see the Changelog file that
// came with your tarball or type make ChangeLog if you are building from git
@@ -253,6 +254,7 @@
double descent; // max depth below baseline
int refCnt;
GBool ok;
+ GBool hasToUnicode;
};
//------------------------------------------------------------------------