<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<font face="FreeSerif">The problem with OSIS failure to display lemmatization
unless Strong's is also selected is because of line 73 in
osisxhtml.cpp:<br>
<br>
70 void processLemma(bool suspendTextPassThru, XMLTag
&tag, SWBuf &buf) {<br>
71 const char *attrib;<br>
72 const char *val;<br>
73 if ((attrib = tag.getAttribute("lemma"))) {<br>
74 int count =
tag.getAttributePartCount("lemma", ' ');<br>
<br>
Picking on the first word of Tisch Matt.1.1 with diatheke -f xhtml
in gdb, when Strong's is selected with lemmatization (-o nl), then
"tag" is found in processLemma() as<br>
"w lemma=\"strong:G976 lemma.Strong:βιβλος
lemma.ANLEX:βιβλος\" wn=\"001\""<br>
whereas lemmatization alone (-o l) gives "tag" showing<br>
"w savlm=\"strong:G976 lemma.Strong:βιβλος
lemma.ANLEX:βιβλος\"" <br>
<br>
Note "lemma" -vs- "savlm". For -o l, line 70 causes the entire
effort to be skipped.<br>
<br>
I've attached a somewhat ham-handed patch which handles the needed
case, but it has the fatal side effect that it produces
lemmatization even when neither strong's nor lemmatization is
wanted.<br>
<br>
Could someone who deals adequately with filter flow take a further
look and offer a working solution?<br>
</font>
</body>
</html>