I committed an exclusion filter that suppresses all existing findbug
warnings (966). I generated it from the findbugs xml report. Then I
moved all NM_CONFUSING out of the automatically generated block.

When you get a new findbugs warning, you may
1. resolve it;
2. leave it unresolved, and add it to the exclusion filter, with
comment;
3. leave it unsolved, and add it to the automatic block of the
exclusion filter.

I note that we have not accepted findbugs as a tool of the project. I
think that not all developers are a fan of the tool.

The main template of the xslt script that generated the exclusion
filter is as follows:

<xsl:template match='//BugInstance'>
  <xsl:variable name="type" select="@type"/>
  <xsl:for-each select="Class">
    <Match>
      <Class name="{...@classname}"/>
      <xsl:choose>
        <xsl:when test="$type='EQ_DOESNT_OVERRIDE_EQUALS'">
          <xsl:text>&#xA;</xsl:text>
          <xsl:comment> Listing the method 'equals' does not work </xsl:comment>
        </xsl:when>
        <xsl:when test="$type='MF_CLASS_MASKS_FIELD'">
          <xsl:text>&#xA;</xsl:text>
          <xsl:comment> Listing the field does not work; this makes the filter 
apply to all masked fields </xsl:comment>
        </xsl:when>
        <xsl:when test="following-sibling::Method">
          <Method name="{following-sibling::Method[1]/@name}"/>
        </xsl:when>
        <xsl:when test="following-sibling::Field">
          <Field name="{following-sibling::Field[1]/@name}"/>
        </xsl:when>
        <xsl:otherwise>
          <xsl:text>&#xA;</xsl:text>
          <xsl:comment>Neither method nor field</xsl:comment>
        </xsl:otherwise>
      </xsl:choose>
      <Bug pattern="{$type}"/>
    </Match>
  </xsl:for-each>
</xsl:template>

Simon

On Wed, Nov 17, 2010 at 07:24:43PM +0100, Simon Pepping wrote:
> I will take care of this. I am now squashing the easier
> findbugs-reported bugs/problems. Then I will add the remaining ones to
> the exclusion file, such that those that are really excluded can be
> distinguished from those which have not yet been examined.

Reply via email to