Chaṭṭha Saṅgāyana Tipiṭaka (CST)
As known by most Pāli students, the most complete Pāli text collection in digital form is of the CST4 contributed by Vipassana Research Institute. The most widely used version is possibly the corpus bundled with the CST4 program. Now when errors are accumulatedly detected, the collection has its online corrected version as Tipitaka.org XML.
Textual correction report
To keep pace with the recent corrections, I also apply the changes to the collections bundled with Pāli Platform 3 (PP3), particularly the CST-Restructured and Grammatical text collection. For the former CST4 collection, now the texts are changed to the better versions. Furthermore, when I edited the files, several more errors were detected and fixed. As a result, the collections bundled with PP3 (from version RC1 onward), as well as the online collections maintained by me here, are now possibly the least erroneous ones.
To make my corrections visible, I made a report showing each instance of editing. By its complexity, I made the report in spreadsheets as shown in the links below.
Tipiṭaka Correction Report
Here are some explanations of the spreadsheets:
- The Main table shows erroneous instances found in CSTR collection against the Tipitaka-XML (corrected up to August 2024). The treatments are also applied to the collection of CST-Kit.
- The Gram table shows erroneous instances found in grammatical texts both in PP3 and this website.
- The erroneous instances uncommon to the both collections above are not reported.
- The Add1 table shows erroneous instances undetected formerly in the Tipitaka-XML. These corrections are applied to the current CST4 collection, as well as the CSTR and Gram collection.
- The Add2 table shows possibly erroneous instances with triple letters. Some obviously need to be fixed. Some need to be analyzed further.
- The Add3 table shows some unusual sequences undetected formerly. Only some obvious cases are fixed.
- The Unusual table shows possibly mistreated instances in the former corrections of Tipitaka-XML.
Character analyzing report
Before a text collection is utilized by PP3, it must be closely analyzed to find potential defects and the best way to approach the text, such as indexing strategy. When the CST4 data were used, they underwent several treatments before the collection was safely enough to be bundled.
This character analysis is not done by those who maintain the Tipitaka-XML collection. When I turn to use this better collection, I have to apply the treatments again (in fact I have done this several times with the collections based on CSCD).
To make these erroneous points visible to the text maintainers, I therefore add this report here below.
Tipiṭaka-XML Analyzing Report
Some may be curious how to do character analysis. You can see the program and other tools in Tipiṭaka-Kit
Manually nti fixing report
When I make a Pāli text collection, I normally prefer n’ti over ’nti. That is because I mainly work with Roman texts and the ’nti version of them makes some information lost when the texts are tokenized (the accusative marker is a marked one).
Mostly we can fix the nti programmatically. See the actual program for this corpus in Tipiṭaka-Kit. Unfortunately, it is far from perfect. There are some points that need manually fixes. Here is the report of the manually nti fixes applied to the CSTR and Gram collection, and to the alternative (corrected) CST4 collection.
Manually ‘nti’ Fixing Report
Buddha Jayanthi Tripitaka (BJT)
The tipitaka.lk’s BJT texts hosted in GRETIL are really messy. A better digital version of that collection is maintained by Path Nirvana Foundation at this Github. So, I decide to remove the GRETIL BJT corpus from PP3 and use this better version instead (available in PP3 RC1 onward).
When I worked on this corpus as I always do for every text collection I use, I also found several errors myself, reported in the links below. Because this corpus uses CC-BY-ND license, I leave these errors unedited.
BJT Pāli Correction Report
This corpus is easier to deal with than the CST4 XML, and it has undergone a similar treatments. You can see my programs used here at Tipiṭaka-Kit. And please go through its README first.