Parse .strings file with Python

Parse .strings file with Python

up vote1down votefavorite

I'm trying to write a small Python script to parse the .strings file in my iPhone application project and determine which keys might not be in use. I'm, also doing some string matching to filter out some of the results. This is where my problems start :). If I try something like for file_line in strings_file: if 'search_keyword' in file_line: ... the search keyword will often not match, even though if I print every file line in the same for I seem to be reading the text correctly and my search keywords appear. The problem is these .strings files are in some binary format. Does anyone know of a proper way to parse these files? iphone python objective-c core-foundation share|improve this question edited Jan 25 '10 at 13:16 asked Jan 25 '10 at 12:49 Mihai Damian 2,51631736 80% accept rate

Can you post an example .strings file? – Tim Pietzcker Jan 25 '10 at 12:57 They are pretty standard in format. I have lines like: "Keyword" = "Value"; or "Keyword" = 0.0; The files are correct as far as they're concerned. I use them from Objective-C and I have no problems there. It's just that there I use some framework provided class to open them which I probably can't use from Python – Mihai Damian Jan 25 '10 at 13:04 So how do you open the file? What does printing the repr() of a line you think should match show you? – Thomas Wouters Jan 25 '10 at 13:05 1   I open it with open(filename, 'r'). Also tried 'U'. 'b' is not supported on Mac OS – Mihai Damian Jan 25 '10 at 13:11 It just occurred to me I might be able to use PyObjC to use the Objective-C code that reads these files. I'm still open to simpler solutions though – Mihai Damian Jan 25 '10 at 13:14 show 1 more comment feedback

3 Answers

activeoldestvotes

up vote1down voteaccepted	Use correct encoding to open the .strings-file and in your source code. According to documentationthe encoding of your file could be utf-16. # -*- coding: utf-8 -*- import codecs for line in codecs.open(u'your_file.strings', encoding='utf-16'): if u'keyword' in line: # process line share|improve this answer answered Jan 25 '10 at 13:27 J.F. Sebastian 61.9k1096173
	Thanks, this works perfectly. Using your solution I realized some of my .strings were encoded utf-8 and most were utf-16. This probably explains why the matching worked only in some of the files. – Mihai Damian Jan 25 '10 at 13:55 feedback

up vote2down vote

No experience with those .strings files, but here is the reason why you don't find matches: strings_file.read() returns a string with the full content of the file. Iterating over a string iterates over single characters, i.e. in your for loop, file_line isn't a line, it's always just one single character (a string of length 1), which obviously can't contain a multi-character search word. share|improve this answer answered Jan 25 '10 at 12:57 balpha♦ 12.9k55177

Indeed. Use for line in strings_file: instead. – Thomas Wouters Jan 25 '10 at 13:01 Ah, you are right, I was using simply for file_line in strings_file:, without the read(); will edit in a moment – Mihai Damian Jan 25 '10 at 13:02 @MihaiD: for line in file will only work (as expected) on plain text files, which (as you say) the.strings files are not. – balpha♦ Jan 25 '10 at 13:05 feedback

up vote0down vote

It sounds like the stings file was saved as data. If python can't read it as is you can convert it to a plain text file in Objective-c. Just: (1) read the strings file into a file with the proper encoding. (2) Convert to dictionary (3) write dictionary to another file. So: NSString *strings=[NSString stringWithContentsOfFile:filePath encoding:NSUTF16StringEncoding error:&error]; NSDictionary *dict=[strings propertyList]; [dict writeToFile:anotherFilePath atomically:NO]; share|improve this answer edited Jan 25 '10 at 14:01 answered Jan 25 '10 at 13:33 TechZen 42.3k74379

Yes, that's one way to go. Will keep in mind if all else fails. – Mihai Damian Jan 25 '10 at 13:47 feedback

Your Answer