How do you search in resource files (*.properties) in Eclipse for string containing non-ASCII characters?
EDIT: Currently I use * in place of those special chars, but I'd prefer Eclipse to handle this for me: so it would either search for '\u00E1' in raw files when I enter 'á', or it might translate the files first and then just search for 'á'.
My apologize for not being specific enough when asking.
In Eclipse, you can use Search -> File Search . In the Search dialog, check the Regular expression option. Then enter this pattern in the Containing text: field to find non-ASCII characters:
[\u007f-\uffff]
(the square brackets are part of the pattern). Enter the File name patterns
*.properties
and then select which resources to search (selected resource, workspace, working set, etc.) and click OK
See also the Pattern javadoc for how to express such regular expressions.
Personally, I search them from the command line using grep, but you can search them in Eclipse by using a question mark in a non-regular-expression search, which should match any character. You can also use a period in a regular expression search.
The search dialog allows you to search for strings in *.resources files in the workspace.
Go to Search -> File. Enable the Regular Expressions checkbox - this also content assist to choose the regular expression according to your needs. In the file name patterns, give *.properties and then, Go :)
Related
I saw a SO post that says you can search using regex or an actual literal text on it to search multiline texts. But what if you want to (quickly) search two or three of words within a specified lines of text content?
For example, what if you want to search for multiline text area that contains "ruby" and "regex" (assuming you want to know where you took a note on your txt (or markdown or rich text format) file. you may want to search for "how to use regex in ruby" or "the ruby regex tutorial", right? )
Now you can use a simple (but redundant) regex like ruby(.*\n)+regex|regex(.*\n)+ruby. But to me it doesn't look beautiful. For three or more words, this kind of regex workaround increases its redundancy exponentially also, not good.
So is there a smarter way to do this? Thanks.
I want to, using the DocX library [https://docx.codeplex.com/], convert a .docx document to use a different font. Does anybody know how to do that? The samples projects are very spare, and the documentation is nonexistent.
I find, too, that often there are extraneous spaces in documents, and I want to iterate over all these until there are never two contiguous spaces. I can do this in a loop, I guess, replacing " " (2 spaces) with " " (1 space) until " " (2 spaces) is no longer found.
However, I also want to remove superfluous line breaks that sometimes occur when copying-and-pasting text into a document. I can do it "manually" (in Libre Office, not sure how it's done in MS Word), as I got an answer to this question:
(select "Regular Expressions" and then replace "$" (without the quotes) with a space)
...but how programmatically, with DocX?
Additionally, in some cases I want to ADD line breaks/"paragraph returns" where there are legitimate line breaks between the end of one paragraph and the start of another, but no extra line to separate them visually. According to this:
...I can add a paragraph/line break to a legitimate line break by searching for "$" and replacing that with "\n\n"
This does work, too (manually, in Libre Office); but again...how to do this with the DocX library?
It appears that not all of this is possible with the current version of the DocX library you are using. If it is not exposed in documentation, the functions might as well not exist, and you should not be using undocumented features.
There is a much more mature library available, however, called the "Open XML SDK", that can do everything you need.
The correct way to change a font, regardless of whether you are doing it with the document editor, or you are writing a program to manipulate these files, is to change the appropriate text's style attribute, or changing the definition of style in use.
You should never, ever, ever, ever directly change the font of any text. Personally, I think that the 'font type' and 'font size' menus should be removed entirely from word/libreoffice/etc, and only be accessible inside a 'change style properties' dialog; the only reason to directly apply a font is if you are actually providing an example of particular typeface under discussion!
See How to: Replace the styles parts in a word processing document (Open XML SDK) from the MSDN documentation for a description of the way that works.
To search and replace text, the applicable MSDN page is How to: Search and replace text in a document part (Open XML SDK). For specifically replacing multiple spaces with a single space, there are numerous results on Google that should all work to at least some degree.
We have some XML file that contains some invalid character, and the program says neither which file it is, nor which line number or character offset. It would be a few seconds work to fix the problem if I could just search for exactly that character, but I cannot find how to express a Unicode character in the file search (or at least I assume so, since the search returns nothing).
Neither 0x1e nor \u001e seem to match anything.
[EDIT] I mean, I can still change the code, and eventually find which file it is by catching the Exception, and using some kind of script/tool to find where exactly the character is, but I do believe it should be possible to search with Unicode in Eclipse, and that is what I am asking in this question.
It may be a problem with the character encoding.
As you're going to need to perform a global / site-wide search to find the , you'll probably need to set the global text file encoding:
Preferences -> Workspace -> Text file encoding
This option may be under the 'General' section in Eclipse, depending on your setup and installed plugins etc.
Ensure that the encoding is set to UTF-8.
You will also need to escape the unicode character sequences, like so:
\u2665
(which I see you have tried)
As we're enabling our (English) application to be localized, we're replaced all in-line strings with NSLocalizedString() calls. Since all of our English strings are all there in-line with the code, e.g. NSLocalizedString(#"OK, #"OK button in a message box"), is there any reason we need the English version of Localizable.strings? When we try removing the strings from the English Localizable.strings, the program seems to work fine. Just wanted to double check if there was some side-effect to not having that around. Thanks, alex
One of the main points of using the NSLocalizedString() macro is so that your programming code can be parsed with the genstrings command-line tool to generate the corresponding Localizable.strings file(s) (see Resource Programming Guide: About the String-Loading Macros and Using the Genstrings Tool to Create Strings Files).
That Localizable.strings file then serves as a starting point for your translators, to use to translate to another language. Without that file to work with, your translators would basically need access to your source code in order to see all the strings you want to use (which kind of defeats the purpose).
Yes, your English version works fine right now, since if a localized version of the string you try to get in code –– for example, NSLocalizedString(#"OK", #"") –– cannot be found in a .strings file, it simply uses the #"OK" string that you passed in.
Another reason why you should likely be keeping the English Localizable.strings is that you should generally try to avoid using high-ASCII characters in your code, but should use the full range of available characters in your actual user interface. For example, you may not want to put the following characters in your code, but would want to use them in your user interface:
… (horizontal ellipsis) (U+2026)
“ (left double quotation mark) (U+201C)
” (right double quotation mark) (U+201D)
‘ (left single quotation mark) (U+2018)
’ (right single quotation mark) (U+2019)
So in code, you'd do something like this:
NSLocalizedString(#"Add Bookmark...", #"")
and then in your .strings file (which is UTF16, so this is fine):
"Add Bookmark..." = "Add Bookmark…";
It's not recommended to use the words as keys, if you want to change that Ok for something else and you have any other Localizable.strings, you will have to edit every single of them to update the Okkey.
Surely your translators will want your Localizable.strings file to work from. And want it to be ALL the strings.
(It is true that the system will fall back to the key, but relying on that seems like bad practice. And you'll find it more reliable to use proper punctuation in a Unicode file, which source code seldom is.)
Mainly I want to search for all lines that contain XXX and YYY on the same line in Eclipse.
What would be the correct search expression for that?
This regex should comply with your request:
(XXX.*YYY|YYY.*XXX)
Used under File Search, checking Regular expression.
Use XXX*YYY in file search option.
This will mean that any characters can come in between XXX and YYY.