topic Searching a string in various directories in SAS Programming

Searching a string in various directories

Carl_1234 — Tue, 12 Mar 2024 18:34:17 GMT

This program searches for a specific string in various directories on my computer. After running it asks to input a string and also to select a specific directory. I just want this program to search all directories at once so that I don't have to select and search them one by one.

%global fnobs dir ext;
%let ext = SAS;  /*file extension*/
%let string = %UPCASE(&SEARCH_STRING);
%PUT &STRING;
proc datasets nolist;
delete all;
run;
%MACRO SETDIR;
%IF &DBSYS = CboTestEditValidation %THEN
%DO;
			%LET DBSYS=CBO_2324_TEST_EDIT_VALIDATION;
			%LET DIR=\\edw-fs02\CAE_Data\CBO\&Year\Programs;
%END;

%ELSE %IF &DBSYS = CBOSurveyReports %then %do;
	%let dbsys = CBO-Survey-Reports-Programs;
	%LET DIR=\\edw-fs02\CAE_Data\CBO\CBO Survey Reports\&Year;
%end;

%ELSE %IF &DBSYS = WDISPrograms %then %do;
	%let dbsys = WDIS;
	%LET DIR=\\edw-fs02\CAE_Data\WDIS\PRODUCTION\Survey Reports\&Year;
%end;

%ELSE %IF &DBSYS = SECCTEAppendixFF %then %do;
	%let dbsys = Appendix;
	%LET DIR=\\edw-fs02\CAE_Data\APPENDICES_WDIS_SECCTE\SECCTE\FF_appendixff_xls;
%end;

%ELSE %IF &DBSYS = SECCTEAppendixY %then %do;
	%let dbsys = Appendix;
	%LET DIR=\\edw-fs02\CAE_Data\APPENDICES_WDIS_SECCTE\SECCTE\Y_appendixy_xls;
%end;

%ELSE %IF &DBSYS = WDISAppendixH %then %do;
	%let dbsys = Appendix;
	%LET DIR=\\edw-fs02\CAE_Data\APPENDICES_WDIS_SECCTE\WDIS\H_appendixh_xls;
%end;

%ELSE %IF &DBSYS = WDISAppendixO %then %do;
	%let dbsys = Appendix;
	%LET DIR=\\edw-fs02\CAE_Data\APPENDICES_WDIS_SECCTE\WDIS\O_appendixo_xls;
%end;

%ELSE %IF &DBSYS = WDISAppendixP %then %do;
	%let dbsys = Appendix;
	%LET DIR=\\edw-fs02\CAE_Data\APPENDICES_WDIS_SECCTE\WDIS\P_appendixp_xls;
%end;

%ELSE %IF &DBSYS = WDISAppendixQ %then %do;
	%let dbsys = Appendix;
	%LET DIR=\\edw-fs02\CAE_Data\APPENDICES_WDIS_SECCTE\WDIS\Q_appendixq_xls;
%end;

%ELSE %IF &DBSYS = SECCTEWDISAppendixZ %then %do;
	%let dbsys = Appendix;
	%LET DIR=\\edw-fs02\CAE_Data\APPENDICES_WDIS_SECCTE\WDIS_SECCTE_AppendixZ_2324;
%end;

%ELSE %IF &DBSYS = BuildCIPSupportFiles %then %do;
	%let dbsys = Appendix;
	%LET DIR="\\edw-fs02\CAE_Data\CIPS\2023-2024\Build CIP and Support Files";
%end;

%ELSE %IF &DBSYS = 2324WDISReferenceTables %then %do;
	%let dbsys = Appendix;
	%LET DIR=\\edw-fs02\CAE_Data\CIPS\2023-2024\2324 WDIS REFERENCE TABLES;
%end;


/*%ELSE %IF &DBSYS = CBO2324SurveyReports %then %do;
	%let dbsys = CBO-2324-Survey-Reports-Programs;
	%LET DIR=\\edw-fs02\CAE_Data\CBO\CBO Survey Reports\2324;
%end;

%ELSE %IF &DBSYS = CBO2425SurveyReports %then %do;
	%let dbsys = CBO-2425-Survey-Reports-Programs;
	%LET DIR=\\edw-fs02\CAE_Data\CBO\CBO Survey Reports\2425;
%end;

%ELSE %IF &DBSYS = CBO2526SurveyReports %then %do;
	%let dbsys = CBO-2526-Survey-Reports-Programs;
	%LET DIR=\\edw-fs02\CAE_Data\CBO\CBO Survey Reports\2526;
%end;

%ELSE %IF &DBSYS = WDISPrograms2324 %then %do;
	%let dbsys = WDIS-2324;
	%LET DIR=\\edw-fs02\CAE_Data\WDIS\PRODUCTION\Survey Reports\2324;
%end;

%ELSE %IF &DBSYS = WDISPrograms2425 %then %do;
	%let dbsys = WDIS-2425;
	%LET DIR=\\edw-fs02\CAE_Data\WDIS\PRODUCTION\Survey Reports\2425;
%end;

%ELSE %IF &DBSYS = WDISprograms2526 %then %do;
	%let dbsys = WDIS-2526;
	%LET DIR=\\edw-fs02\CAE_Data\WDIS\PRODUCTION\Survey Reports\2526;
%end;*/
%ELSE %LET DIR=\\edw-fs02\CAE_Data\WDIS\PRODUCTION\Survey Reports\2324;

%PUT &DIR;
%MEND SETDIR;
%SETDIR;
RUN;
%MACRO SETOBS (DSET);

	DATA _NULL_;
		DSID=OPEN("&DSET");

		IF (DSID = 0) THEN DO;
			PUT 'FILE NOT OPENED';
		END;
		ELSE DO;
			PUT 'DSET DATA SET HAS BEEN OPENED';
			ISOBS=ATTRN(DSID,'NOBS');

			IF ISOBS GT -1 THEN DO;
				CALL SYMPUT ("FNOBS",RIGHT(PUT(ISOBS,8.)));
			END;
		END;
	RUN;
%MEND SETOBS;
%macro list_members;
	%global dnum;

	data tibco_list(KEEP=filename ext);
		format fref $char8. filename $char256. filetype $char100. CCNUM 8. CCNUMX $char2. colname $char50. sub $char1. trm 1. year $char4. dbtype $char3.
			ext $char3.;
		rc = filename(fref,"&DIR");

		if rc = 0 then
			do;
				did = dopen(fref);
				rc = filename(fref);
			end;
		else
			do;
				length msg $200.;
				msg = sysmsg();
				put msg=;
				did = .;
			end;

		if did <= 0 then
			putlog 'ERR' 'OR: Unable to open directory.';
		dnum = dnum(did);
		call symput("dnum",dnum);
		fcount = 0;
		%PUT &DNUM;

		do j = 1 to dnum;
			filename = dread(did, j);
			putlog filename=;
			fid = mopen(did, filename);
			ext=upcase(scan(filename,-1,'.'));
			NUMDOTS = count(filename,'.');
			putlog numdots=;

			IF ext IN ('SAS', 'egp') THEN
				output tibco_list;
		end;

		rc = dclose(did);
	run;

%mend list_members;

%list_members;
RUN;

%macro readraw;

	%SETOBS(tibco_list);

	DATA tibco_list;
	SET tibco_list NOBS=NOBS;
	RETAIN SEQ;
	SEQ=SUM(SEQ,1);
	RUN;

	%let rc=%sysfunc(filename(fileref,%superq(dir)));
	%let did=%sysfunc(dopen(&fileref));

	%do dmem=1 %to &fnobs;

	   	DATA _NULL_;
		SET tibco_list;
		WHERE SEQ = &dmem;
		CALL SYMPUT("memname",trim(filename));
		RUN;

		*--this is telling the code that we are only interested in files that have an extension of SAS;
		%if %qupcase(%qscan(%superq(memname),-1,.))=&ext. %then
			%do;
				*--depending on the file type you are looking at, you may need another method to read in the file contents;
				data temp;
					length program $3600.;
					infile "&dir.\&memname." truncover;
					input @01 line $char2500.;
					line_num=_n_;
					program="&dir\&memname.";

					if index(upcase(line), "&STRING") > 0 then
						do;
							line_num=_n_;
							program="&dir\&memname.";
							output;
						end;
				run;

				proc append data=temp base=all;
				run;

			%end;
	%end;

	%let dsid = %sysfunc(open(all));

	*-- If the data set exists, then grab the number of observations and variables;
	*-- then close the data seth;
	%if &dsid %then
		%do;
			%let nobs =%sysfunc(attrn(&dsid,nobs));
			%let rc = %sysfunc(close(&dsid));
			%put &nobs;

			%if &nobs > 0 %then
				%do;
					proc print data=all noobs;
						title "Looking for  &string.";
						title2 " &string found in the following programs";
					run;
				%end;
			%else %put NOTE: &string not found in &dir.;
		%end;

	%let didc=%sysfunc(dclose(%nrbquote(&did)));
	%let rc=%sysfunc(filename(fileref));
%mend readraw;

%readraw;
run;

Re: Searching a string in various directories

SASJedi — Tue, 12 Mar 2024 19:18:06 GMT

For a head start, I have a ready-made macro (%findfiles) that can traverse directories and produce either a list of files with a specified extension in the log
or write them to a SAS dataset.
SYNTAX:
%FindFiles(dir,<ext,dsn,sub>)
dir=fully qualified directory path
ext=Space delimited list of file extensions (Optional, default is ALL)
dsn=name of data set to store filenames (Optional, otherwise writes to log.)
sub=look in subfolders? (Y|N default is Y)

Examples:
%FindFiles(c:\temp, csv)
%FindFiles(\\server\folder\, xls xlsx xlsm, work.myfiles)
%FindFiles(s:/workshop,sas,work.pgm_files,N)

This macro depends on three utility macros ( %exist, %fileattribs, and %translate). You can download the files from my GitHub sas-macros repository, or use this SAS code to download them:

%let path=<your custom macro storage location>;
filename maccode "&path/exists.sas";
proc http url="https://github.com/SASJedi/sas-macros/raw/master/exist.sas" out=maccode;
run;

filename maccode2 "&path/fileattribs.sas";
proc http url="https://github.com/SASJedi/sas-macros/raw/master/fileattribs.sas" out=maccode2;
run;

filename maccode3 "&path/translate.sas";
proc http url="https://github.com/SASJedi/sas-macros/raw/master/translate.sas" out=maccode3;
run;

filename maccode4 "&path/findfiles.sas";
proc http url="https://github.com/SASJedi/sas-macros/raw/master/findfiles.sas" out=maccode4;
run;

Examine the files you downloaded to see how the code works and to ensure there's nothing nefarious going on there, if you wish.

/* After examining the code you downloaded, run the macro programs to compile the macors */
%include maccode;
%include maccode2;
%include maccode3;
%include maccode4;

filename maccode clear;
filename maccode2 clear;
filename maccode3 clear;
filename maccode4 clear;

/*Get syntax help in the log for any macro using ? as the only parameter value:*/
%findfiles(?)

You can then use the dataset you created to drive the rest of your process...

For example, I ran this on my computer:

%FindFiles(D:/Courses/PG1V2, xls xlsx xlsm, work.myfiles)
proc print noobs; run;

And this was the result:

Item	Path	Filename	Size	CRDate	CRTime	ModDate	ModTime
1	D:/Courses/PG1V2/data	class.xlsx	10,000	06/12/2023	7:51:04	12/08/2017	12:13:00
2	D:/Courses/PG1V2/data	eu_sport_trade.xlsx	234,100	06/12/2023	7:51:04	12/19/2017	14:48:36
3	D:/Courses/PG1V2/data	np_info.xlsx	7,234,320	06/12/2023	7:51:04	01/30/2018	13:26:40
4	D:/Courses/PG1V2/data	storm.xlsx	6,859,276	06/12/2023	7:51:04	12/05/2019	12:51:42

Hope that helps.

Re: Searching a string in various directories

ballardw — Tue, 12 Mar 2024 20:19:29 GMT

Which operating system?

DOS (and WIndows still supports Dos commands) FINDSTR

UNIX and derivative operating systems the GREP command does this

I would start with OS commands instead of writing from scratch as many of those are already set up to do directory and subdirectory searches.