My current function under test is supposed to handle extraction of archived files. Zip, unzip — it sounds so easy! But often times, things aren’t as easy as they sound. Inevitably, I came across some unexpected problems though while I was testing:
- My previous version of 7zip (4.57) changed the special characters to underscores. This was remedied by an upgrade to 7zip 4.62.
- I wanted to come up with a “corrupted” file. To achieve that, I just opened the zip file with my trusty text editor and tampered with the contents.
- I used PowerArchiver to create a spanned archive. In one case, I included only one zip file from the set. In another, I included all zip files from the set. In both cases, no decompress error was raised. The folders turned out empty though.
- I used 3 different tools to create archives of the a folder containing special characters (Chinese characters, ☼, !@#$%^&-_=) — Windows’ send to compressed (zipped) folder, PowerArchiver and 7zip. I kept getting decompress errors for the zip files created by PowerArchiver.
- Using the mac mini, I created an archive with a folder containing special characters. During decompression though, “繁馬陳門診從務數” became “τ╣üΘª¼ΘÖ│ΘûÇΦ¿║σ╛₧σïÖµò╕” and “la niña” became “la nin╠âa”.
- Using a linux machine, I created an archive. By default (either that, or “.tar.gz” was the last format chosen), the archive created had a .tar.gz extension. We’re supposed to support gzip files, but not tar files. But since linux’s default archival would involve tar files, it looks like we’d have to support it. My dev’s quite quick — the fix for this is already included in the next build!