1
0
mirror of https://github.com/sasjs/core.git synced 2025-12-10 22:14:35 +00:00

fix: supporting special missings in BETWEEN and IN operators

Impacts mp_filtercheck.sas.  Tests added.  Closes #279
This commit is contained in:
Allan Bowe
2022-07-07 22:47:04 +00:00
parent b377b83442
commit 6e32eb3bd6
4 changed files with 81 additions and 28 deletions

39
all.sas
View File

@@ -5883,7 +5883,8 @@ data &outds;
/*length GROUP_LOGIC SUBGROUP_LOGIC $3 SUBGROUP_ID 8 VARIABLE_NM $32
OPERATOR_NM $10 RAW_VALUE $4000;*/
set &inds;
length reason_cd $4032 vtype $1 vnum dsid 8;
length reason_cd $4032 vtype $1 vnum dsid 8 tmp $4000;
drop tmp;
/* quick check to ensure column exists */
if upcase(VARIABLE_NM) not in
@@ -5959,18 +5960,32 @@ data &outds;
end;
/* special logic */
if OPERATOR_NM='BETWEEN' then raw_value1=tranwrd(raw_value,' AND ','');
else if OPERATOR_NM in ('IN','NOT IN') then do;
if substr(raw_value,1,1) ne '('
or substr(cats(reverse(raw_value)),1,1) ne ')'
then do;
REASON_CD='Missing start/end bracket in RAW_VALUE';
putlog REASON_CD= OPERATOR_NM= raw_value= raw_value1= ;
call symputx('reason_cd',reason_cd,'l');
call symputx('nobs',_n_,'l');
output;
if OPERATOR_NM in ('IN','NOT IN','BETWEEN') then do;
if OPERATOR_NM='BETWEEN' then raw_value1=tranwrd(raw_value,' AND ',',');
else do;
if substr(raw_value,1,1) ne '('
or substr(cats(reverse(raw_value)),1,1) ne ')'
then do;
REASON_CD='Missing start/end bracket in RAW_VALUE';
putlog REASON_CD= OPERATOR_NM= raw_value= raw_value1= ;
call symputx('reason_cd',reason_cd,'l');
call symputx('nobs',_n_,'l');
output;
end;
else raw_value1=substr(raw_value,2,max(length(raw_value)-2,0));
/* we now have a comma seperated list of values */
if vtype='N' then do i=1 to countc(raw_value1, ',')+1;
tmp=scan(raw_value1,i,',');
if cats(tmp) ne '.' and input(tmp, ?? 8.) eq . then do;
REASON_CD='Non Numeric value provided';
putlog REASON_CD= OPERATOR_NM= raw_value= raw_value1= ;
call symputx('reason_cd',reason_cd,'l');
call symputx('nobs',_n_,'l');
output;
end;
return;
end;
end;
else raw_value1=substr(raw_value,2,max(length(raw_value)-2,0));
end;
else raw_value1=raw_value;

View File

@@ -92,7 +92,8 @@ data &outds;
/*length GROUP_LOGIC SUBGROUP_LOGIC $3 SUBGROUP_ID 8 VARIABLE_NM $32
OPERATOR_NM $10 RAW_VALUE $4000;*/
set &inds;
length reason_cd $4032 vtype $1 vnum dsid 8;
length reason_cd $4032 vtype $1 vnum dsid 8 tmp $4000;
drop tmp;
/* quick check to ensure column exists */
if upcase(VARIABLE_NM) not in
@@ -168,18 +169,32 @@ data &outds;
end;
/* special logic */
if OPERATOR_NM='BETWEEN' then raw_value1=tranwrd(raw_value,' AND ','');
else if OPERATOR_NM in ('IN','NOT IN') then do;
if substr(raw_value,1,1) ne '('
or substr(cats(reverse(raw_value)),1,1) ne ')'
then do;
REASON_CD='Missing start/end bracket in RAW_VALUE';
putlog REASON_CD= OPERATOR_NM= raw_value= raw_value1= ;
call symputx('reason_cd',reason_cd,'l');
call symputx('nobs',_n_,'l');
output;
if OPERATOR_NM in ('IN','NOT IN','BETWEEN') then do;
if OPERATOR_NM='BETWEEN' then raw_value1=tranwrd(raw_value,' AND ',',');
else do;
if substr(raw_value,1,1) ne '('
or substr(cats(reverse(raw_value)),1,1) ne ')'
then do;
REASON_CD='Missing start/end bracket in RAW_VALUE';
putlog REASON_CD= OPERATOR_NM= raw_value= raw_value1= ;
call symputx('reason_cd',reason_cd,'l');
call symputx('nobs',_n_,'l');
output;
end;
else raw_value1=substr(raw_value,2,max(length(raw_value)-2,0));
/* we now have a comma seperated list of values */
if vtype='N' then do i=1 to countc(raw_value1, ',')+1;
tmp=scan(raw_value1,i,',');
if cats(tmp) ne '.' and input(tmp, ?? 8.) eq . then do;
REASON_CD='Non Numeric value provided';
putlog REASON_CD= OPERATOR_NM= raw_value= raw_value1= ;
call symputx('reason_cd',reason_cd,'l');
call symputx('nobs',_n_,'l');
output;
end;
return;
end;
end;
else raw_value1=substr(raw_value,2,max(length(raw_value)-2,0));
end;
else raw_value1=raw_value;

View File

@@ -67,7 +67,7 @@
},
{
"name": "server",
"serverUrl": "",
"serverUrl": "https://sas.4gl.io",
"serverType": "SASJS",
"httpsAgentOptions": {
"allowInsecureRequests": false
@@ -107,4 +107,4 @@
"contextName": "SAS Job Execution compute context"
}
]
}
}

View File

@@ -38,7 +38,7 @@ William,M,15,66.5,112
;;;;
run;
/* valid filter conditions */
/* VALID filter conditions */
data work.inds;
infile datalines4 dsd;
input GROUP_LOGIC:$3. SUBGROUP_LOGIC:$3. SUBGROUP_ID:8. VARIABLE_NM:$32.
@@ -51,6 +51,7 @@ AND,OR,2,Weight,>=,77.7
AND,OR,2,Weight,NE,77.7
AND,AND,1,age,=,.A
AND,AND,1,height,<,.B
AND,AND,1,age,IN,"(.a,.b,.)"
;;;;
run;
@@ -115,6 +116,28 @@ run;
outds=work.test_results
)
/* invalid IN value */
data work.inds;
infile datalines4 dsd;
input GROUP_LOGIC:$3. SUBGROUP_LOGIC:$3. SUBGROUP_ID:8. VARIABLE_NM:$32.
OPERATOR_NM:$10. RAW_VALUE:$4000.;
datalines4;
AND,OR,2,age,IN,"(.,.a,X)"
;;;;
run;
%mp_filtercheck(work.inds,
targetds=work.class,
outds=work.badrecords,
abort=NO
)
%let syscc=0;
%mp_assertdsobs(work.badrecords,
desc=Invalid IN value,
test=HASOBS,
outds=work.test_results
)
/* Code injection - column name */
data work.inds;
infile datalines4 dsd;
@@ -163,7 +186,7 @@ run;
data work.inds;
infile datalines4 dsd;
input GROUP_LOGIC:$3. SUBGROUP_LOGIC:$3. SUBGROUP_ID:8. VARIABLE_NM:$32.
OPERATOR_NM:$10. RAW_VALUE:8;
OPERATOR_NM:$10. RAW_VALUE:8.;
datalines4;
AND,AND,1,age,=,0
;;;;