Santhosh has uploaded a new change for review.
https://gerrit.wikimedia.org/r/123519
Change subject: Segmentation: Fix the wrong usage of section types
......................................................................
Segmentation: Fix the wrong usage of section types
Eventhough the sectionTypes array was introduced, openTag handler was
not using it by mistake
Change-Id: I1a5c49102b545a369d48ca74319826e393dd5a9a
---
M segmentation/languages/CXParser.js
M tests/segmentation/SegmentationTests.json
2 files changed, 5 insertions(+), 6 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/mediawiki/services/cxserver
refs/changes/19/123519/1
diff --git a/segmentation/languages/CXParser.js
b/segmentation/languages/CXParser.js
index d3499d5..a57502f 100644
--- a/segmentation/languages/CXParser.js
+++ b/segmentation/languages/CXParser.js
@@ -146,8 +146,7 @@
* @param {Object} tag
*/
CXParser.prototype.onopentag = function ( tag ) {
- var attrName,
- section = /[ph1-6]|figure|figcaption|ul|div/;
+ var attrName;
if ( this.sectionTypes.indexOf( tag.name ) >= 0 ) {
if ( this.inSentence ) {
@@ -172,7 +171,7 @@
}
// Sections
- if ( tag.name.match( section ) ) {
+ if ( this.sectionTypes.indexOf( tag.name ) >= 0 ) {
if ( !tag.attributes.id ) {
this.print( ' id="' + ( this.segmentCount++ ) + '"' );
}
@@ -182,7 +181,7 @@
this.print( '>' );
// Start the first segment of the section
- if ( tag.name.match( section ) ) {
+ if ( this.sectionTypes.indexOf( tag.name ) >= 0 ) {
this.print( this.startSentence() );
}
};
diff --git a/tests/segmentation/SegmentationTests.json
b/tests/segmentation/SegmentationTests.json
index 4de98c8..b150841 100644
--- a/tests/segmentation/SegmentationTests.json
+++ b/tests/segmentation/SegmentationTests.json
@@ -34,8 +34,8 @@
},
{
"desc": "References can appear after period. Example:
Hydrogen is a gas. [1] It is ...",
- "source": "<p>Sentence one. <a href=\"#\">reference</a>
Starts with reference</p>",
- "result": "<p id=\"0\"><span class=\"cx-segment\"
data-segmentid=\"1\">Sentence one. </span><span class=\"cx-segment\"
data-segmentid=\"2\"><a class=\"cx-link\" data-linkid=\"3\"
href=\"#\">reference</a> Starts with reference</span></p>"
+ "source": "<p>Sentence one. <span
class=\"reference\"><a href=\"#\">reference</a></span> Starts with
reference</p>",
+ "result": "<p id=\"0\"><span class=\"cx-segment\"
data-segmentid=\"1\">Sentence one. </span><span class=\"cx-segment\"
data-segmentid=\"2\"><span class=\"reference\"><a class=\"cx-link\"
data-linkid=\"3\" href=\"#\">reference</a></span> Starts with
reference</span></p>"
}
],
"hi": [
--
To view, visit https://gerrit.wikimedia.org/r/123519
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: I1a5c49102b545a369d48ca74319826e393dd5a9a
Gerrit-PatchSet: 1
Gerrit-Project: mediawiki/services/cxserver
Gerrit-Branch: master
Gerrit-Owner: Santhosh <santhosh.thottin...@gmail.com>
_______________________________________________
MediaWiki-commits mailing list
MediaWiki-commits@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
