Baoding-FenShuaJiang
/
Green_GoodERP18_FSJ_Standard
보기 1
좋아요 0
위키 捐赠
gooderp18绿色标准版
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
브렌치: master
브랜치 태그
${ item.name }
Create branch ${ searchTerm }
from 'master'
${ noResults }
Green_GoodERP18_FSJ_Standard/PostgreSQL/doc/postgresql/html/regress-variant.html

77 lines
7.0KB
Raw 고유링크 Blame 히스토리 

  1. <?xml version="1.0" encoding="UTF-8" standalone="no"?>

  2. <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>32.3. Variant Comparison Files</title><link rel="stylesheet" type="text/css" href="stylesheet.css" /><link rev="made" href="pgsql-docs@lists.postgresql.org" /><meta name="generator" content="DocBook XSL Stylesheets V1.79.1" /><link rel="prev" href="regress-evaluation.html" title="32.2. Test Evaluation" /><link rel="next" href="regress-tap.html" title="32.4. TAP Tests" /></head><body><div xmlns="http://www.w3.org/TR/xhtml1/transitional" class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">32.3. Variant Comparison Files</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="regress-evaluation.html" title="32.2. Test Evaluation">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="regress.html" title="Chapter 32. Regression Tests">Up</a></td><th width="60%" align="center">Chapter 32. Regression Tests</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 12.4 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="regress-tap.html" title="32.4. TAP Tests">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="REGRESS-VARIANT"><div class="titlepage"><div><div><h2 class="title" style="clear: both">32.3. Variant Comparison Files</h2></div></div></div><p>

  3.     Since some of the tests inherently produce environment-dependent

  4.     results, we have provided ways to specify alternate <span class="quote">“<span class="quote">expected</span>”</span>

  5.     result files.  Each regression test can have several comparison files

  6.     showing possible results on different platforms.  There are two

  7.     independent mechanisms for determining which comparison file is used

  8.     for each test.

  9.    </p><p>

  10.     The first mechanism allows comparison files to be selected for

  11.     specific platforms.  There is a mapping file,

  12.     <code class="filename">src/test/regress/resultmap</code>, that defines

  13.     which comparison file to use for each platform.

  14.     To eliminate bogus test <span class="quote">“<span class="quote">failures</span>”</span> for a particular platform,

  15.     you first choose or make a variant result file, and then add a line to the

  16.     <code class="filename">resultmap</code> file.

  17.    </p><p>

  18.     Each line in the mapping file is of the form

  19. </p><pre class="synopsis">

  20. testname:output:platformpattern=comparisonfilename

  21. </pre><p>

  22.     The test name is just the name of the particular regression test

  23.     module. The output value indicates which output file to check. For the

  24.     standard regression tests, this is always <code class="literal">out</code>. The

  25.     value corresponds to the file extension of the output file.

  26.     The platform pattern is a pattern in the style of the Unix

  27.     tool <code class="command">expr</code> (that is, a regular expression with an implicit

  28.     <code class="literal">^</code> anchor at the start).  It is matched against the

  29.     platform name as printed by <code class="command">config.guess</code>.

  30.     The comparison file name is the base name of the substitute result

  31.     comparison file.

  32.    </p><p>

  33.     For example: some systems lack a working <code class="literal">strtof</code> function,

  34.     for which our workaround causes rounding errors in the

  35.     <code class="filename">float4</code> regression test.

  36.     Therefore, we provide a variant comparison file,

  37.     <code class="filename">float4-misrounded-input.out</code>, which includes

  38.     the results to be expected on these systems.  To silence the bogus

  39.     <span class="quote">“<span class="quote">failure</span>”</span> message on <span class="systemitem">HP-UX 10</span>

  40.     platforms, <code class="filename">resultmap</code> includes:

  41. </p><pre class="programlisting">

  42. float4:out:hppa.*-hp-hpux10.*=float4-misrounded-input.out

  43. </pre><p>

  44.     which will trigger on any machine where the output of

  45.     <code class="command">config.guess</code> matches <code class="literal">hppa.*-hp-hpux10.*</code>.

  46.     Other lines in <code class="filename">resultmap</code> select the variant comparison

  47.     file for other platforms where it's appropriate.

  48.    </p><p>

  49.     The second selection mechanism for variant comparison files is

  50.     much more automatic: it simply uses the <span class="quote">“<span class="quote">best match</span>”</span> among

  51.     several supplied comparison files.  The regression test driver

  52.     script considers both the standard comparison file for a test,

  53.     <code class="literal"><em class="replaceable"><code>testname</code></em>.out</code>, and variant files named

  54.     <code class="literal"><em class="replaceable"><code>testname</code></em>_<em class="replaceable"><code>digit</code></em>.out</code>

  55.     (where the <em class="replaceable"><code>digit</code></em> is any single digit

  56.     <code class="literal">0</code>-<code class="literal">9</code>).  If any such file is an exact match,

  57.     the test is considered to pass; otherwise, the one that generates

  58.     the shortest diff is used to create the failure report.  (If

  59.     <code class="filename">resultmap</code> includes an entry for the particular

  60.     test, then the base <em class="replaceable"><code>testname</code></em> is the substitute

  61.     name given in <code class="filename">resultmap</code>.)

  62.    </p><p>

  63.     For example, for the <code class="literal">char</code> test, the comparison file

  64.     <code class="filename">char.out</code> contains results that are expected

  65.     in the <code class="literal">C</code> and <code class="literal">POSIX</code> locales, while

  66.     the file <code class="filename">char_1.out</code> contains results sorted as

  67.     they appear in many other locales.

  68.    </p><p>

  69.     The best-match mechanism was devised to cope with locale-dependent

  70.     results, but it can be used in any situation where the test results

  71.     cannot be predicted easily from the platform name alone.  A limitation of

  72.     this mechanism is that the test driver cannot tell which variant is

  73.     actually <span class="quote">“<span class="quote">correct</span>”</span> for the current environment; it will just pick

  74.     the variant that seems to work best.  Therefore it is safest to use this

  75.     mechanism only for variant results that you are willing to consider

  76.     equally valid in all contexts.

  77.    </p></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="regress-evaluation.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="regress.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="regress-tap.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">32.2. Test Evaluation </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> 32.4. TAP Tests</td></tr></table></div></body></html>
上海开阖软件有限公司 沪ICP备12045867号-1