Reported Terms in Proper Case

1library(yamaa)23yamaa_spec("spec.yaml")$run()$output

1import yamaa23yamaa.spec("spec.yaml").run().output

1# Reference solution for the yamaa benchmark adam-adae-proper-case-terms (R track).2#3# Written from the benchmark's full prompt and its inputs alone, as an agent4# would write it. Harbor's oracle agent runs it to check that the prompt can5# be solved and that the grader scores a correct answer 1.67ae <- read.csv(8  "/app/input/ae.csv",9  colClasses = "character",10  na.strings = ""11)1213# The first character, and each character that directly follows one of the14# delimiters, in upper case; every other letter in lower case. A term with no15# value is "Not Reported".16propcase <- function(terms, delimiters) {17  separators <- strsplit(delimiters, "")[[1]]18  vapply(terms, function(term) {19    if (is.na(term)) {20      return("Not Reported")21    }22    characters <- strsplit(term, "")[[1]]23    previous <- c("", head(characters, -1))24    raise <- seq_along(characters) == 1 | previous %in% separators25    paste(ifelse(raise, toupper(characters), tolower(characters)), collapse = "")26  }, character(1), USE.NAMES = FALSE)27}2829adae <- data.frame(30  STUDYID = ae$STUDYID,31  USUBJID = ae$USUBJID,32  AESEQ = as.integer(ae$AESEQ),33  AETERM = ae$AETERM,34  # SAS PROPCASE's default delimiters, then the ones that replace them.35  AETERMPC = propcase(ae$AETERM, " /-(.\t"),36  AETERMPB = propcase(ae$AETERM, " /-(.,[")37)3839dir.create("/app/output", showWarnings = FALSE)40write.csv(adae, "/app/output/adae.csv", row.names = FALSE, na = "")

1# Reference solution for the yamaa benchmark adam-adae-proper-case-terms (Python track).2#3# Written from the benchmark's full prompt and its inputs alone, as an agent4# would write it. Harbor's oracle agent runs it to check that the prompt can5# be solved and that the grader scores a correct answer 1.67import csv89INPUT = "/app/input/ae.csv"10OUTPUT = "/app/output/adae.csv"1112COLUMNS = ["STUDYID", "USUBJID", "AESEQ", "AETERM", "AETERMPC", "AETERMPB"]1314# SAS PROPCASE's default delimiters, and the ones that replace them for15# AETERMPB.16DEFAULT = " /-(.\t"17COMMA_AND_BRACKET = " /-(.,["181920def propcase(text, delimiters):21    """The first character, and each character that directly follows one of22    the delimiters, in upper case; every other letter in lower case."""23    return "".join(24        character.upper()25        if index == 0 or text[index - 1] in delimiters26        else character.lower()27        for index, character in enumerate(text)28    )293031def main():32    with open(INPUT, newline="", encoding="utf-8") as handle:33        rows = list(csv.DictReader(handle))3435    out_rows = []36    for row in rows:37        term = row.get("AETERM") or ""38        out_rows.append(39            {40                "STUDYID": row["STUDYID"],41                "USUBJID": row["USUBJID"],42                "AESEQ": row["AESEQ"],43                "AETERM": term,44                "AETERMPC": propcase(term, DEFAULT) if term else "Not Reported",45                "AETERMPB": (46                    propcase(term, COMMA_AND_BRACKET) if term else "Not Reported"47                ),48            }49        )5051    with open(OUTPUT, "w", newline="", encoding="utf-8") as handle:52        writer = csv.DictWriter(handle, fieldnames=COLUMNS)53        writer.writeheader()54        writer.writerows(out_rows)555657if __name__ == "__main__":58    main()

1/* opensas reference solution for the yamaa benchmark adam-adae-proper-case-terms.2   Reads the benchmark inputs and derives the requested output.3   Execute with the existing opensas runtime: opensas /app/output/result.sas. */45data ae;6  length STUDYID USUBJID AETERM $1024;7  infile "/app/input/ae.csv" dsd dlm="," firstobs=2 truncover lrecl=32767;8  input STUDYID : $char1024. USUBJID : $char1024. AESEQ AETERM : $char1024.;9run;1011/* Proper case as PROPCASE writes it: the first character, and each one that12   directly follows a delimiter, in upper case, and every other letter in13   lower case. Each term is rebuilt one character at a time, a blank14   included. AETERMPC takes the default delimiters, the tab among them, and15   AETERMPB the ones that replace them. The previous character is compared16   with each delimiter: INDEX finds no blank excerpt in this release. */17data result;18  length AETERMPC AETERMPB $1024 CH PREV PC PB $1;19  set ae;20  if missing(AETERM) then do;21    AETERMPC = "Not Reported";22    AETERMPB = "Not Reported";23  end;24  else do;25    do I = 1 to lengthn(AETERM);26      CH = substr(AETERM, I, 1);27      if I = 1 then do;28        AETERMPC = upcase(CH);29        AETERMPB = upcase(CH);30      end;31      else do;32        PREV = substr(AETERM, I - 1, 1);33        if PREV in (" ", "/", "-", "(", ".", '09'x) then PC = upcase(CH);34        else PC = lowcase(CH);35        if PREV in (" ", "/", "-", "(", ".", ",", "[") then PB = upcase(CH);36        else PB = lowcase(CH);37        AETERMPC = substr(AETERMPC, 1, I - 1) || PC;38        AETERMPB = substr(AETERMPB, 1, I - 1) || PB;39      end;40    end;41  end;42run;4344proc sql;45  create table final as select STUDYID, USUBJID, AESEQ, AETERM, AETERMPC, AETERMPB from result;46quit;47proc export data=final outfile="/app/output/adae.csv" dbms=csv replace; run;