ZuMult instance for TGDP

[berndmoos]

Taking notes here, may turn them into separate issues later.

• SearchIndexer is constructing filepath in a Windows specific way: SearchIndex searchIndex = index.createIndex(inputDirectories, INDEX_PATH + "\\"+ INDEX_NAME, MTAS_CONFIG_FILE_PATH + "\\" +MTAS_CONFIG_FILE_NAME);
• NetBeans configuration file nbactions.xml has hard coded path to Java executable: <exec.executable>C:\\Program Files\\Java\\jdk-17\\bin\\java.exe</exec.executable>

[berndmoos]

• The static initialisation for COMAFileSystem takes very long to make an index of IDs. There are about 20.000 such IDs in TGDP. Indexing takes about 15 minutes. Different solution? Write the index to file?

[berndmoos]

• webservices/SearchService: getKWICviaULR-- should this be getKWICviaURL?

corrected

[berndmoos]

• search-index-path in Configuration.xml wants a file separator (backslash on Windows) at the end while other file paths do not

file separator deleted, code in DGD2searcher.java and COMASearcher.java modified accordingly

[berndmoos]

Results of day 1 :-)

[berndmoos]

To do on day 2:

• find out about how to index/query for xml:lang tags (token level)
• index the entire corpus, might need a COMAIndexer wrapper for this (because the transcripts are in many subfolders)?
• fix audio references

[berndmoos]

• It is quite unclear what to do with transcript metadata. We never needed it for the AGD backend, but it contains crucial information (in COMA) for TGDP ==> #148

[berndmoos]

[berndmoos]

[berndmoos]

This is now all in https://github.com/berndmoos/zumulttgdp