Hyphens and spaces omitted in scraped words

[dkalantzi]

Hello,

Thank you for this very useful resource. I've noticed two potential issues with the words in the scraped data:

• Hyphens Omission: Hyphens from the original dictionary entries seem to be missing. For example, the first word in the txt is aalugalog, but the entry in the dictionary is aalug-alog.
• Spaces Omission: Spaces between words in phrases from the original dictionary also seem to be missing, causing phrases to be scraped as single words. For example, patay na Bulan is scraped as pataynaBulan, “patay na hayop” is scraped as pataynahayop.

Kind regards,
Dimitra

[nikitimi]

Hi @dkalantzi, I'm building an application like Duolingo, but for the native dialects here in the Philippines.
I've came across this tagalog web scraper by @raymelon , so here is the response regarding your issue:

You can changed the Regular Expression in line 161 of collect_tagalog.py

tagalog-dictionary-scraper/collect_tagalog.py

Line 161 in 37b6c0b

all_words.append(re.compile('[^a-zA-Z]').sub('', word.next.next))

with:

all_words.append(re.compile('[ ]?\(+?[\s\w\d\W]+\)').sub('', word.next.next).lower())

• The '[ ]' checks for whitespaces, you can also write \s.
• The following flags: \w is for words, \d is for decimals, \W is for non-words like '&'.
• Lastly all words will be converted to lowercase with lower() method.

Check RegExr's RegEx reference for more information about Regular Expressions.

Here is the result of scraped words with the modified RegEx above.
tagalog_dict.txt

You can adjust the regex to get the result that is in your liking.