source: other-projects/maori-lang-detection/lib/commons-csv-1.8.jar@ 36373

Last change on this file since 36373 was 33940, checked in by ak19, 4 years ago
  1. In order to make it easier to do the manual work of inspecting 260 web pages, there's now a new class ManualURLInspection.java that will automate the part where I have to paste a URL into mongo db and press Find to locate the matching web page's record and then view the record. The new class takes the input CSV file and for each of the 260 web page URLs of the random sample, it gets the full text from mongodb and displays it on the console and let's me enter Y/N/? for whether I think the overall page is in MRI or not. Then the updated records are written into a tmp file, as well as any remainiing ones I haven't gone through yet. If I'm satisfied, I'll replace the original sample csv file with the new one and can continue from where I left off later. Responds to Ctrl-D for end of input. Some handling of Ctrl-C: each processed line is written out, but user is instructed to manually append the unprocessed records to the end of the temp output file before replacing original csv input file with output file to continue working later. 2. Needed to update commons-csv jar file to have access to additional method. 3. Needed a new utility method in MongodbQueryer. 4. Minor changes to SummaryTool.java. 5. Processed first 190 of 260 random sample web urls using my helpful new class ManualURLInspection.java. The first 15 URLs were done totally manually, the remainder with my new class. The new class made the work very much faster and far less tedious.
  • Property svn:mime-type set to application/octet-stream
File size: 47.9 KB

HTML preview not available, since no preview renderer could handle it.Try downloading the file instead.

Note: See TracBrowser for help on using the repository browser.