Using Grep to find/replace

I'm trying to find out how to use GREP in find/replace to chage the formatting of some text that comes in from a spreadsheet.
I worked out the GREP query "~b(\d\d)~b", which finds a paragraph return, followed by two digits, followed by another paragraph return
and then it is replaced by "\t $1~b", which is a tab, the two found digits and a para return.
What i need to do is to amend the query to find ANY number of digits, (which may be comma delimited: eg 23, 36, 48 ,50), and then replace with a tab + found text.
I suppose what I'm looking for is a way for the query to find "any text between two paragraph returns, no matter what tthe length", but I don't know how to do this.All the Wildcard options seem to find just one exampler (one digit, one character etc)

And you came so far
The operators for repeat are ? (zero or once), * (zero or more) and + (once or more). You can also specify exact numbers: {at least,up to}.
All of these operators are "greedy" by default -- they will match as much as possible. To match as least as possible (which I'm sure you'll come up against, sooner or later), add another ? after the repeat expression.
So this will find one digit, then optionally another (which will always be included):
\d\d?
and this one digit, then zero or as much as ten million million zillion:
\d\d*
which is functionally the same as
\d+
And this will find between 3 and 8 digits but will forced to use the shortest possible match:
\d{3,8}?
That said: A quick & dirty solution for your actual problem is to find any amount of digits, spaces, and comma's:
~b[\d, ]+~b
(we need the plus here because otherwise it would also match an empty line). The [..] brackets an Inclusive list --- it will match any of the single codes inside.
A more complicated but 'neater' way is to search very specifically only for number, comma, space, number sequences -- it's neater because that way malformed lines (comma without a space) will be skipped!
(It also introduces another code -- the parentheses operators. Look them up in a good GREP reference --lost of people are enthousiastic about Peter Kahrel's O'Reilly title, because it's about using GREP in InDesign.)
~b\d+(, \d+)*~b

Similar Messages

Using grep to find one match per line

Hi,
How do you get grep to return a single line even when multiple matches are found on the same line?
I tried the (?-m) and (?-s) without success.
eg. searching for the string "sin" in the following line should return only one result not 2.
SG SIN Singapore
I don't want the search to stop after a single match, it should continue on successive lines
Thanks,
SB

Sure. Let GREP find an entire line at once: >^.*sin.*$ But it won't work, I'm afraid. You insist that looking for "sin" i will find a match in "SG SIN Singapore"; and not one but two. It won't. Try it. <pre> __ (..) _\</pre> Got that? So use the 'case insensitive' switch >(?i)^.*sin.*$ or (perhaps better, in this case), as you seem to search for the three letter codes, use >^.*\<SIN\>.*$

How to use GREP to find one particular word with automatic hyphen?

I know I have in my text a word with automatic hyphen on the end of line, say Wo-rd.
How can I find it?
Thanks.

I think you might do better to edit the hyphenation in the Dictionary.
One reason words hyphenate incorrectly is the wrong language has been applied. Both spell checking and hyphenation rules are governed by the language selected for the text, and language is a character-level attribute so that you can select inidvidual words and assign a different language to them if you like.
If that's the probelm here, you can look for the word using find/change (without worying about the hyphen) and change the language for all instances.
In either case you will be far better off fixing the problem on a global scale in the document than you will be finding a particular instance of that word at the end of a line. Any editing you do in the future may cause the problem to reappear somewhere else.

Finished script: Use grep find/change to fill in a supplied table of contents

This script is now complete, and has been the subject of most of my previous posts. Just in case anyone wanted to know what the finished script ended as, here it is.
Thanks so much to all. A lot of really helpful folks on this board are very responsible for the success of this task. This script is to be one of hopefully many in the creation of our records. But it's a huge leap forward. Thanks again to everyone that helped.
Cheers,
~Nate
Task:
Automatically find town names in listings, and fill in table of contents template on page 2 accordingly.
Example of page 2 toc, initially:
Example of a page of content. The town names are what need to be referenced on the TOC:
Example of page 2 toc once script is finished:
Because of the need to include the transaction dates on the TOC (comes as a provided, tagged-text file), a simple Indesign-generated TOC can't be used alone.
This script uses an Indesign-generated TOC that's on a master page called "T-tocGen" ... It then uses grep search and replaces to grab the needed information, and insert it into the page 2 TOC.
The script will update a generated TOC and then search for an instance of a page number, and town name. The generated toc lists all included towns in the following format:
(line start)## tab townName(line end)
In Grep, this would be (please note, extra \ for \d and \t ... javascript needs that for some reason):
^\\d+\\t(.*)$
After the script gets the info it needs from a found instance of the above, it replaces that line with "---", to prevent that line from being picked up once again.
The script with then place the needed page number in it's rightful place on page 2, replacing the XX.
A while loop is used to repeat the above process until there are no longer any instances of "^\\d+\\t(.*)$" present.
Not every town runs every issue, so once the script is done, it removes all remaining instance of "XX" on the page 2 TOC.
FINAL CODE:
TOC replace
This script will use grep find/change methods to apply page numbers in
tocGen to the XX's on page2TOC.
// define the text frame of generated TOC
    var tocGenFrame = document.masterSpreads.item("T-tocGen").pages.item(0).textFrames.item(0);
// udpate generated TOC ... store contents in tocGenStuff
    var tocGenStuff = updateTOCGen();
// set variable for while loop
var okGo = "1";
// while okGo isn't 0
while(okGo.length!=0)
// get town info from tocGen
getCurrentTown();
// replace XX's with tocGen info
replaceTown();
// grep find ... any remaining towns with page numbers in tocGen?
app.findGrepPreferences = app.changeGrepPreferences = null;
app.findGrepPreferences.findWhat = "^\\d+\\t(.*)$";
// set current value of okGo ... with any instances of above grep find in tocGen
okGo = tocGenFrame.findGrep();
// grep find/change all leftover XXs in page2TOC
app.findGrepPreferences = app.changeGrepPreferences = null;
app.findGrepPreferences.findWhat = "^XX\\t";
app.changeGrepPreferences.changeTo = "\\t";
app.activeDocument.changeGrep();
// clear grep prefs
app.findGrepPreferences = app.changeGrepPreferences = null;
// functions                  //
function getCurrentTown()
// grep options
app.findChangeGrepOptions.includeLockedLayersForFind = true;
app.findChangeGrepOptions.includeLockedStoriesForFind = true;
app.findChangeGrepOptions.includeHiddenLayers = true;
app.findChangeGrepOptions.includeMasterPages = true;
app.findChangeGrepOptions.includeFootnotes = true;
// grep find: startLine anyDigits tab anyCharacters endLine
      app.findGrepPreferences = app.changeGrepPreferences = null;
      app.findGrepPreferences.findWhat = "^\\d+\\t(.*)$";
// get grep find results
currentGen = tocGenFrame.findGrep();
// store grep results content into currentLine
currentLine = currentGen[0].contents;
// match to get array of grep found items
currentMatch = currentGen[0].contents.match("^\\d+\\t(.*)$");
// second found item is town name, store as currentTown
currentTown = currentMatch[1];
// change current line to --- now that data has been grabbed
// this is because loop will continue as long as the above grep find yields a result
       app.findGrepPreferences.findWhat = "^\\d+\\t"+currentTown+"$";
              app.changeGrepPreferences.changeTo = "---";
            tocGenFrame.changeGrep();
function replaceTown()
app.findChangeGrepOptions.includeLockedLayersForFind = true;
app.findChangeGrepOptions.includeLockedStoriesForFind = true;
app.findChangeGrepOptions.includeHiddenLayers = true;
app.findChangeGrepOptions.includeMasterPages = true;
app.findChangeGrepOptions.includeFootnotes = true;
// find: XX currentTown .... replace with: currentLine
    app.findGrepPreferences = app.changeGrepPreferences = null;
    app.findGrepPreferences.findWhat = "^XX\\t"+currentTown+" \\(";
    app.changeGrepPreferences.changeTo = currentLine+" \(";
app.activeDocument.changeGrep();
function updateTOCGen()
//set vars ... toc text frame, toc master pag
    var tocGen = document.masterSpreads.item("T-tocGen").pages.item(0).textFrames.item(0);
    var tocGenPage = document.masterSpreads.item("T-tocGen").pages.item(0);
//SELECT the text frame generatedTOC on the master TOC
    tocGen.select();
//Update Table of Contents by script menu action:
    app.scriptMenuActions.itemByID(71442).invoke();
//Deselect selection of text frame holding your TOC:
    app.select(null);
//store contents of toc text frame in variable
    var tocGenText = tocGen.contents;
//return contents of tocGen
    return tocGenText;

Thanks for the reply.
You are correct but the problem is there are three rows, One row is 100% black, the second is 60% black and the third is 40% black. I want to change the black to blue, the 60% black to an orange and the 40% black to a light shaded blue. In the find/change option you can select the tint you want to find and replace but yea.. does work on table cells.. oddly enough.

Using Grep on multiline textframe

Hello
I am using GREP to find text and replace in scripting it but It seems to not work on multiline text. Ex.:
My texte
My text2
I use ^.({8})(My text2){1}) to replace the 'My text2' but it doesnt work. Is there any way to make safe text subtituions with grep?

If you have a return between the two lines, the GREP needs to take that into account... (i.e. \r)
Harbs

[GREP] – find/replace only first comma in paragraph

Is it possible to find and replace only the first comma in paragraph with certain paragraph style with end of paragraph character using GREP search? Or in other words – to split paragraph into two paragraphs by replacing the first comma in paragraph with end of paragraph character?
Claudius

That's an interesting problem, and I think it takes at least two steps.
The expression (^[^,]+),\s* will find all text from the beginning of a paragraph up to the first comma, which might be followed by whitespace and allows you to use the $1 operator in the change to field to save the text up to the comma, but replace the comma and following whitespace with anything you like. However, as soon as you replace that with a paragraph break you've created a new paragraph to search and you end up breaking at every comma.
Instead you need to insert a stand-in symbol of some sort with a unique character style applied (so it won't get picked up any other way), for instance change to $1% and apply the style named "break" which does no formatting but flags the text for the next search.
Now search for % with the "break" style and replace with the paragragh break (and no style).
You should probably now search for .+ with the break style and change to nothing (or $0 for found text), and remove the style, just to be clean.
There's still one problem, though, and that's that all the new paragraphs start with a lower-case character unless the first word after the comma happened to be capitalized. You could now search for ^. to find the first character in each paragraph and use the change format to apply a character style that assigns all caps, but that's really an ugly way to work.
All of this will fall apart, too, I think if you have nested or GREP styles affecting the beginning text or the first character in any paragraph that you wind up with.

Ignore Line break in find text using grep

Hi everyone!
I need to find the text in document using grep.
find text for :
Xereptatiuria que alique volo eium qui dolupid ut
voluptatiam earum saestorepel iuscit im quas et modisimodit.
The above sentence cannot having line breaks. But document having multiple line breaks.
so, please to give a tip to find the text using grep.
I am not excepting this type of result by following.
Xereptatiuriaque\n alique volo\neium qui dolupid ut\r voluptatiam earum\n saestorepel iuscit im quas et\n modisimodit. ------------this is not.
Any another way to find (i.e)., ignore the line break
simply like this,
(?s:Xereptatiuria que alique volo eium qui dolupid ut
voluptatiam earum saestorepel iuscit im quas et modisimodit.) --------- the line break wherever it is found,ignore line break in single command.
Thanks by,
John Peter.

johnp45247251,
what do you really want to do?
Like pkahrel in your other thread Find Grep And Ignore the line break said:
pkahrel schrieb:
You're a moving target: you change your question in each post. Do yourself a favour and go and read up on GREP.
johnp45247251, one question:
Could it be possible, that you doesn't understand correctly, how Grep really works?
<edit by pixxxel schubser>
Furthermore your example text is different to your screenshot. Your text is:
»Xereptatiuriaque\n alique volo\neium qui dolupid ut\r voluptatiam earum\n saestorepel iuscit im quas et\n modisimodit.«
And your screenshot shows:
»Xereptatiuriaque\n alique volo\n eium qui dolupid ut \rvoluptatiam earum\n saestorepel iuscit im quas et\n modisimodit.«
Normally the text should look like this:
»Xereptatiuriaque\nalique volo\neium qui dolupid ut\rvoluptatiam earum\nsaestorepel iuscit im quas et\nmodisimodit.«
Please explain what do you really want to do and exactly if additional spaces are exists or not!
Could it be:
you have a text with paragraphs and with many line breaks in that paragraph and there are no spaces exists before or after your line breaks. And is your destination to remove all the line breaks from your text?
</edit by pixxxel schubser>
Then the way is not to find the text completly and to ignore the line breaks – the way is to find the line breaks and replace with a space, e.g. like that:
find:
\n
replace with:
\s
(change all)
Regards
pixxxel schubser

Grep find/replace to clean-up texts

Hi,
I want to use The ID GREP find/replace to clean-up a imported text with at every line a carriage/return (CR).
(space)CR -> (space)
a_wordCR -> aword(space)
CRCR -> CR
How can I perform this with Grep expressions ?
Thanks

In the first and 3rd replaces, you want a CR replaced with nothing; in the 2nd one, with a space. So it's not possbile to do this with one single "magic" search and replace -- Grep is not magic
This query will take care of the returns:
(?<=[\r ])\r
(replace with nothing). It finds returns with at their left either a return or a single space.
This one
(?<=\w)\r
(replace with space) will change a return after a word character to a space. "Word character" is a bit undefined -- it appears to be at least A-Z (including accented characters), 0-9, and Greek/Cyrillic characters. In case of a problem, you could try
(?<=[[:alpha:]])\r
instead -- forcing 'alphabetic' characters only.

Find & replace part of a string in Numbers using do shell script in AppleScript

Hello,
I would like to set a search-pattern with a wildcard in Applescript to find - for example - the pattern 'Table 1::$*$4' for use in a 'Search & Replace script'
The dollar signs '$' seem to be a bit of problem (refers to fixed values in Numbers & to variables in Shell ...)
Could anyone hand me a solution to this problem?
The end-goal - for now - would be to change the reference to a row-number in a lot of cells (number '4' in the pattern above should finally be replaced by 5, 6, 7, ...)
Thx.

Hi,
Here's how to do that:
try
 tell application "Numbers" to tell front document to tell active sheet
 tell (first table whose selection range's class is range)
 set sr to selection range
 set f to text returned of (display dialog "Find this in selected cells in Numbers " default answer "" with title "Find-Replace Step 1" buttons {"Cancel", "Next"})
 if f = "" then return
 set r to text returned of (display dialog "Replace '" & f & "' with " default answer f with title "Find-Replace Step 2")
 set {f, r} to my escapeForSED(f, r) -- escape some chars, create back reference for sed
 set tc to count cells of sr
 tell sr to repeat with i from 1 to tc
 tell (cell i) to try
 set oVal to formula
 if oVal is not missing value then set value to (my find_replace(oVal, f, r))
 end try
 end repeat
 end tell
 end tell
on error number n
 if n = -128 then return
 display dialog "Did you select cells?" buttons {"cancel"} with title "Oops!"
end try
on find_replace(t, f, r)
 do shell script "/usr/bin/sed 's~" & f & "~" & r & "~g' <<< " & (quoted form of t)
end find_replace
on escapeForSED(f, r)
 set tid to text item delimiters
 set text item delimiters to "*" -- the wildcard
 set tc1 to count (text items of f)
 set tc2 to count (text items of r)
 set text item delimiters to tid
 if (tc1 - tc2) < 0 then
 display alert "The number of wildcard in the replacement string must be equal or less than the number of wildcard in the search string."
 error -128
 end if
 -- escape search string, and create back reference for each wildcard (the wildcard is a dot in sed) --> \$.\$
 set f to do shell script "/usr/bin/sed -e 's/[]~$.^|[]/\\\\&/g;s/\\*/\\\$.\\\$/g' <<<" & quoted form of f
 -- escape the replacement string, Perl replace wildcard by two backslash and an incremented integer, to get the back reference --> \\1 \\2
 return {f, (do shell script "/usr/bin/sed -e 's/[]~$.^|[]/\\\\&/g' | /usr/bin/perl -pe '$n=1;s/\\*/\"\\\\\" . $n++/ge'<<<" & (quoted form of r))}
end escapeForSED
For what you want to do, you must have the wildcard in the same position in both string. --> find "Table 1::$*$3", replace "Table 1::$*$4"
Important, you can use no wildcard in both (the search string and the replacement string) or you can use any wildcard in the search string with no wildcard in the replacement string).
But, the number of wildcard in the replacement string must be equal or less than the number of wildcard in the search string.

Is there a way to use the find/replace tool to select the actual step in the TestStand sequence instead of just listing matching values in the find window?

The purpose of this is to be able to select many steps that are scattered throughout a sequence so I can change the step type one time. I could do this now, but it would take me 30 minutes to click through the sequence and select all of the steps before I'd even change the step type because there are different steps inbetween the ones I want to select. I wouldn't be able to just use the shift key to select multiple steps because of that. I have about 500 steps I need to select, so it would be worth it to find an automated way to do that.
Unless there's a way to change a step type from a numericlimittest to an action step type from the find/replace window???

I ended up using ActiveX calls in TestStand to find the steps I wanted and change the step type to Action. See attached example sequence.
Attachments:
Convert NumericLimitTest to Action step type example.seq ‏9 KB

How can I remove all content between two tags using Find/Replace regular expressions?

This one is driving me bonkers... I'm relatively new to regular expressions, but I'm trying to get Dreamweaver to remove all content between two tags in an XML document. For example, let's say I have the following XML:
<custom>
<![CDATA[Some text
Some more text]]>
</custom>
I'd like to do a Find/Replace that produces:
<custom>
</custom>
In essence, I'd like to strip all of the content between two tags. Ideally, I'd like to know how to strip the CDATA content as well, to return the following:
<custom>
<![CDATA[]]>
</custom>
I'd much appreciate any suggestions on accomplishing this.
Many thanks!

Thanks much for your response. I found David's article to be a little thin with respect to examples using quantifiers in coordination with the wildcard metacharacters; however, I was able to cobble together a working expression through trial and error using the information he presented. For posterity, here’s the solution:
Find:
<custom>[\d\D]*?</custom>
Replace:
<custom>
<![CDATA[]]>
</custom>
I believe this literally translates to:
[] = find anything in this range/character class
\d = find any digit character (i.e. any number)
\D = find any non-digit character (i.e. anything except numbers)
*? = match zero or more times, but as few times as possible (i.e. match multiple characters per instance, but only match one instance at a time, or none at all)
I’m still not sure how to effectively utilize the . wildcard. For example, the following expression will not find content that ends with a number:
<custom>.*?[\D]*?</ custom >
I'm presuming this is because numbers aren't included in the \D metacharacter; however, shouldn't numbers be picked up by the .*? expression?

Why not there a "Find/replace in Grep style" inside the Para style?

Whenever I type a digit in my text, it should be colored red as per style. I do this by grep style inside the para style, but now I need to insert brackets before and after of the digit(s), i realize that there is no replace option in grep style in the para style. Why not it be there a "find/replace" instead "find" only as it now appears?

Ya, this is simple, finding a specific para style with digit and change them, when the book in first pass. But while in the correction pass of the same book, whenever we are inserting more text into the document, there are chances to be unaware of the digit style that, it should be surrounded by brackets, and it happened earlier so I have a thought of it. Again, while paginating a book having more than 350 - 600 pages, and 3 to 4 guys working in it, I think this may work.
Expecting your valuable comment on this.
Thangaraj Mohan.

Find/Replace Using Regular Expressions

Can someone help me with this...I am using Regular expressions to
FIND:
http.*lid=([^&"]*)[^"]*
REPLACE:
$set(\1,ID_id,code)$
So that in the following it will change this:
a href="http://www.test.com/shc/s/home_10153_12605?lid=Search" rilt="Search"
To this:
a href="$set(Search,ID_id,code)$" rilt="Search
Those expressions work in Notepad++ but when i use dreamweaver it just replaces the http... with "$set(\1,ID_id,code)$" and doesnt reference the "search"
Any help?
Thanks

Let me begin by saying I'm a complete idiot with DW's Reg Ex. I use Search Specific Tag whenever possible. See screenshot below.
Try this on your Current Document to see if it works. Then make a back-up copy of site before attempting it on Entire Local Site as you cannot "Undo" this process.
Good luck,
Nancy O.

Use Find-Replace to paste/wrap Element name?

Hi:
For FrameMaker 7.2, structured.
I'm a longtime *unstructured* FrameMaker who is performing cleanup on some structured FrameMaker documents created by my customer.
One of my tasks is to find all Index markers and applying the IndexMarker element (square-cornered) tag to the markers.
I'm currently using the tedious and slow technique of using Find to locate and select the Index tag, then in the Element catalog clicking "IndexMarker" followed by "Wrap".
I've poked around hoping I could find Copy Special / Paste Special choices that would speed this mousing-intensive chore up, but haven't found any such choices that work.
So -- is there some "intelligent" way to Find / Replace my way through this problem? Even if I manually have to click Replace at every instance, it's faster than what I'm doing...
Cheers & thanks,
Riley

Riley,
As much as I appreciate the unsolicited endorsements, I have to clarify that FrameSLT cannot help you with this task. It can do lots of things associated with structural management, but all of its current capabilities are dependent upon existing structure. More specifically, it requires existing structure to find the items for which it intends to manage, which your current markers do not have. So, while it can do things like element wrapping, it wouldn't be able to find the unstructured markers in order to wrap them.
This would be a reasonably simple task for a custom API client or FrameScript script. If you have zillions of these to do, it might be worth your while to explore that option. If not, Gary's suggestion might be your best bet.
Russ
(Owner, West Street Consulting)

Can I use Find & Replace to change the fills and strokes of tables?

I'm working on a catalog that has a fair amount of tables, that all have the same color scheme. I need to change all of these tables to a new color scheme, and I'd much rather find a way to do it with one shot. The tables do not have styles applied to them, so it's not a matter of adjusting the style definitions.
I know you can do a Find & Replace for text or objects in a specific color. Is there a way to do the same for table cells? I've tried but haven't had luck getting it to work.
Thanks!
-Sean

Congratulations are in place -- changing parameter names in the right places is one thing, finding out what parts safely can be copied is step 2 in my book "how to become a successful scripter in 10 easy steps"
You cannot select more than one table in Table mode, but what you can do is select them together with its surrounding text -- just keep on dragging that mouse until you've covered them all. The InDesign Selection is a multi-purpose object (see the way the script has to find out "what is selected"), and one of the things it keeps track of is how many and which tables are "inside" the selection. Then run this slightly amended script:
a = app.selection[0];if (a.tables.length == 0)
alert ("You must be kidding; not a single table inside your selection!");
exit(0);
for (b=0; b<app.selection[0].tables.length; b++)
d = app.selection[0].tables[b]; // shortcut
for (c=0; c<d.cells.length; c++)
if (d.cells[c].fillColor == app.activeDocument.swatches.item("Bloo"))
d.cells[c].fillColor = app.activeDocument.swatches.item("Redd");
// .. Insert more colors here ..
Now the script doesn't have to check anymore in which part of the table you clicked; it only has to iterate over the ones in your selection, and those are "immediately" accessible through the indexes (the square bracketed variables). And the "shortcut" is so I don't have to type "app.selection[0].tables[b].cells[c].fillColor" every time. (And the script runs marginally faster as well, although it's probably still almost immediate.)

Using Grep to find/replace

Similar Messages

Maybe you are looking for