Home • About Us • A-Z Index • Search * Contact Us • Register • Login • Press • Shop

The Open Brand -- Problem Reporting and Interpretations System

Problem Report 0609 Details

Help Show help | Quick Search | Submit a Test Suite Support Request | Click here to view your privileges

This page provides all information on Problem Report 0609.

Report 0609 Actions

    Problem Report Number 0609
    Submitter's Classification Test Suite problem
    State Resolved
    Resolution Rejected (REJ)
    Problem Resolution ID REJ.X.0213
    Raised 1970-01-01 08:00
    Updated 2003-03-13 08:00
    Published null
    Product Standard Commands and Utilities V2 (UNIX 95)
    Certification Program The Open Brand certification program
    Test Suite VSC version 4.1.4
    Test Identification POSIX.upe/ex 447 448
    Problem Summary PG4C.00063 Unexpected shell expansion.
    Problem Text
    POSIX.upe/ex/ex_07.sh - FAILED. Assertion number: tp447 and tp448

    What the test performs :
    =======================
    # Assertion: 447
    # When a regular expression contains the character sequence \e<, then this
    # sequence is matched against the beginning of a word in the lines being
    # matched. (The term word is as defined in POSIX.2 section 5.10.7.2).

    # Strategy:
    #
    # Use the construct \< in a substitute command to replace the beginning
    # of words with a known character (in this case `x'). Verify that the
    # substitution succeeds when words begin with an alphabetic, numeric,
    # or _ character and separated by spaces.
    #
    # Repeat the test with the words separated by non alphanumeric characters
    # from the POSIX locale.

    ----------------------------------------------------------------------
    # Assertion: 448
    # When a regular expression contains the character sequence \e>, then
    # this sequence is matched against the end of a word in the lines being
    # matched. (The term word is as defined in POSIX.2 section 5.10.7.2).

    # Use the construct \> in a substitute command to replace the end
    # of words with a known character (in this case `x'). Verify that the
    # substitution succeeds when words end with an alphabetic, numeric,
    # or _ character and separated by spaces.
    #
    # Repeat the test with the words separated by non alphanumeric characters
    # from the POSIX locale.
    ----------------------------------------------------------------------


    The Assertions #447 and #448 both essentially do similar tests
    except that #447 tests for \e< wherease #448 tests for \e>.

    We will explain in detail the problem with #447 and the same
    shall apply to #448.

    We observe that the test fails for \ and * non-alphanumeric cases.

    What the POSIX Standard states:
    -------------------------------
    Standard : POSIX
    Section : 3.2.2
    Page : 118
    line : 44-47

    "3.2.2 Single Quotes
    Enclosing characters in single quotes (' ') shall preserve
    the literal value of each character within the single
    quotes."


    What the test does is as follows:
    --------------------------------
    The test first assigns non-alphanumeric characters to the
    variable "chars". Among the many non-alphanumric characters,
    the \ and * are the two characters which cause the test to
    fail.

    The following is the assignment to the variable "chars" :

    chars='! @ # $ % ^ & \* ( ) + ~ - = ` { } [ ] | \\ : " ; , . / < >'


    Since the assignment is in single quotes, the variable
    "chars" will retain the literal character sequence that has
    been assigned to it. Consider the following cases:

    1. '\*': The test statement,

    printf "%s\n" "word${c}word" > ex_data_$i

    will generate "word\*word" and not the expected string,
    ^^^^^^^^^^
    "word*word".
    ^^^^^^^^^

    This makes the test fail.

    We understand that the intention of escaping "*" is to
    avoid expansion of "*" to filenames by the shell in the
    "for c in $chars" statement. One possible way of doing
    this will be to use "set -f" in the beginning of the test
    case and initialising "chars" without escaping *.

    2. '\\':

    The test statement,

    printf "%s\n" "word${c}word" > ex_data_$i

    will generate "word\\word" and not the expected string,
    ^^^^^^^^^^

    "word\word".
    ^^^^^^^^^

    This makes the test fail.

    We understand that the intention of escaping the \ is to
    preserve the literal meaning of \ but it is being already
    done by putting \ under single quotes. Hence only \
    should be used instead of \\.


    Based on the above arguments we feel that the tests #447 and
    #448 are in error and should be waivered.

    Test Output
    20|1 1 12812 1 1|Assertion #447 (C): construct \< matches the beginning of a wo
    rd
    520|1 1 12812 1 1|Expected exit code = 0; Received 127
    520|1 1 12812 1 2|char "*" not matched as word delimiter
    520|1 1 12812 1 3|Expected exit code = 0; Received 127
    520|1 1 12812 1 4|char "" not matched as word delimiter
    520|1 1 12812 1 5|2 subtests did not generate expected result
    220|1 1 1 15:12:41|FAIL
    410|1 447 1 15:12:41|IC End

    400|1 448 1 15:12:41|IC Start
    200|1 1 15:12:41|TP Start
    520|1 1 12812 1 1|Assertion #448 (C): construct \> matches the end of a word
    520|1 1 12812 1 1|char "*" not matched as word delimiter
    520|1 1 12812 1 2|char "" not matched as word delimiter
    520|1 1 12812 1 3|2 subtests did not generate expected result
    220|1 1 1 15:29:40|FAIL
    410|1 448 1 15:29:40|IC End
    80|1 0 15:29:41|TC End

    Review Information

    Review Type TSMA Review
    Start Date null
    Completed null
    Status Complete
    Review Recommendation No Resolution Given
    Review Response
    We recommend this request be refused.

    We believe the test is correct.

    Although extra characters are present in some cases during the test,
    they do not result in differences in the results.

    In the for statement, the escaped * character prevents any
    pathname expansion from occurring. Final quote removal does not
    occur, because the escape characters do not appear in the original
    word, $chars. Therefore, the extra backslashes appear in for both
    the * and \ characters.

    The strings resulting from parameter expansion in the for loop,
    "word\*word" and "word\\word", are used in the printf statements for
    the starting input files and an "x" is placed in the appropriate place
    in the expected output files.

    Because the characters are quoted, no pathname expansion is attempted.
    No quote removal occurs because the escape characters do not appear in
    the original word, "word${c}word". Thus, the files contain something
    like "word\*word" or "word\\word".

    The extra characters in the output do not change the results of the test,
    since the definition of word in the Command Descriptions in ex section
    results in the selection of "word" in either case since all other characters
    are treated as delimiters between the words.

    Review Type SA Review
    Start Date null
    Completed null
    Status Complete
    Review Resolution Rejected (REJ)
    Review Conclusion
    This request is refused.

    Problem Reporting System Options:

     

    Back   

Contact the Certification Authority