<div dir="ltr"><div>Kristen, where were you seeing the &quot;No analysis&quot; message? Simply in your LKB window, right?<br></div><div><br></div><div>I just did the following:</div><div><br></div><div>1) looked into the most recent regression test that I added, looked into the &quot;parse&quot; file under home/gold/languagename. Here&#39;s what I see there:</div><div><br></div><div><div>1@1@1@-1@0@-1@0@2@-1@1@1@-1@1@-1@-1@-1@-1@-1@-1@-1@-1@-1@-1@0@0@0@0@0@-1@-1@-1@-1@220636@-1@-1@-1@23-6-2013 14:28:24@0@</div><div>2@1@2@-1@0@-1@0@0@-1@0@0@-1@0@-1@-1@-1@-1@-1@-1@-1@-1@-1@-1@0@0@0@0@0@-1@-1@-1@-1@44636@-1@-1@-1@23-6-2013 14:28:24@post-reduction lexical gap@</div></div><div><br></div><div>So, first item is fine, second one has the error message; incidentally, the second one is ungrammatical, so, in the testsuite, it would start from the star, though that should be completely fine, shouldn&#39;t it? </div><div><br></div><div>The item in the corresponding testsuite looks like this:</div><div><br></div><div>* noun1 noun2 tverb<br></div><div><br></div><div>-- there is a space between the star and &quot;noun1&quot;; but I think that&#39;s how my testsuites have always looked like. Is that correct or should there be no space, or is there any other caveats with the asterisk?</div><div><br></div><div><br></div><div>2) Anyway, then I loaded this grammar (which is stored under gmcs/regression_tests/grammars, until deleted) into the LKB and ran the testsuite with [incr tsdb()]. I don&#39;t think I am seeing anything suspicious, in particular no &quot;No analysis&quot; messages:</div><div><br></div><div><img src="cid:1607560bbdecb971f162" alt="image.png" class="" style="max-width: 100%; opacity: 1;"><br></div><div><br><div class="gmail_quote"><div>Interestingly, the star does not appear to be separated from the sentence by a space, I am assuming the system gets rid of it for this view? Not sure if this is relevant.</div><div><br></div><div>But it must be something going on with the asterisk in my case...</div><div><br></div><div>Olga</div><div dir="ltr">On Wed, Dec 20, 2017 at 10:21 AM Kristen Howell &lt;<a href="mailto:kphowell@uw.edu">kphowell@uw.edu</a>&gt; wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Thanks! It looks like my text editor had added ^M characters at the end of each line, so that&#39;s the unknown word/lexical gap. I&#39;ll update the discourse page with this as well.</div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Dec 20, 2017 at 10:00 AM, Emily M. Bender <span dir="ltr">&lt;<a href="mailto:ebender@uw.edu" target="_blank">ebender@uw.edu</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Hi Kristen,<div><br></div><div>Two things to try:</div><div><br></div><div>(1) Parse the sentence interactively with ACE (rather than the LKB)</div><div>(2) Open the [incr tsdb()] profile and parse the sentence interactively (with the LKB) by double-clicking it</div><span class="m_3599062933696227931HOEnZb"><font color="#888888"><div><br></div><div>Emily</div></font></span></div><div class="gmail_extra"><div><div class="m_3599062933696227931h5"><br><div class="gmail_quote">On Wed, Dec 20, 2017 at 9:57 AM, Woodley Packard <span dir="ltr">&lt;<a href="mailto:sweaglesw@sweaglesw.org" target="_blank">sweaglesw@sweaglesw.org</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="auto"><div>Hi again,</div><div><br></div><div>Dan&#39;s assessment is accurate for the case of the ERG and other grammars that implement unknown word handling.  For smaller grammars however the error will be from cases where the input sentence contains a word that is not in the lexicon (which is what I meant by a lexical gap). More specifically, there is a word for which a stem cannot be found in the lexicon using the existing orthographemic rules.  Since Olga was talking about matrix regression testing I suspect this is her context.</div><span class="m_3599062933696227931m_-3720151325719300329HOEnZb"><font color="#888888"><div><br></div><div>Woodley<br><br><br></div></font></span><div><div class="m_3599062933696227931m_-3720151325719300329h5"><div><br>On Dec 20, 2017, at 9:38 AM, Kristen Howell &lt;<a href="mailto:kphowell@uw.edu" target="_blank">kphowell@uw.edu</a>&gt; wrote:<br><br></div><blockquote type="cite"><div><div dir="ltr">Thanks everyone. I got this parse error a lot with toy grammars that I produced with the grammar matrix. I don&#39;t see lfr.tdl anywhere in those output grammars. Does anyone know what might be the grammar matrix equivalent?</div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Dec 20, 2017 at 12:57 AM, Dan Flickinger <span dir="ltr">&lt;<a href="mailto:danf@stanford.edu" target="_blank">danf@stanford.edu</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">




<div dir="ltr">
<div id="m_3599062933696227931m_-3720151325719300329m_-5267281124168586687m_-3087037634506015501divtagdefaultwrapper" style="font-size:12pt;color:#000000;font-family:Calibri,Helvetica,sans-serif" dir="ltr">
<p style="margin-top:0;margin-bottom:0">Hi Olga,</p>
<p style="margin-top:0;margin-bottom:0"><br>
</p>
<p style="margin-top:0;margin-bottom:0">This error message from ACE occurs when your grammar includes one or more token-mapping (preprocessing) rules which conspire to delete all of the edges from one of the cells in the initial parse chart, so that the parser
 cannot hope to find a covering analysis.  For the ERG, this occasionally happens when I adjust the rules for inserting lexical entries for unknown words, and then try to filter some of those added entries in case I already have a &quot;native&quot; entry for that word. 
 I imagine you must have one or more rules in a file like the ERG&#39;s &quot;lfr.tdl&quot; which discard some unwanted entries before parsing.  Try commenting out these rules and see if the error message for that sentence disappears. You can see the error just by trying
 to parse that one sentence directly with ACE.</p>
<p style="margin-top:0;margin-bottom:0"><br>
</p>
<p style="margin-top:0;margin-bottom:0"> Dan<br>
</p>
<br>
<br>
<div style="color:rgb(0,0,0)">
<hr style="display:inline-block;width:98%">
<div id="m_3599062933696227931m_-3720151325719300329m_-5267281124168586687m_-3087037634506015501divRplyFwdMsg" dir="ltr"><font style="font-size:11pt" color="#000000" face="Calibri, sans-serif"><b>From:</b> <a href="mailto:developers-bounces@emmtee.net" target="_blank">developers-bounces@emmtee.net</a> &lt;<a href="mailto:developers-bounces@emmtee.net" target="_blank">developers-bounces@emmtee.net</a>&gt; on behalf of Olga Zamaraeva &lt;<a href="mailto:olzama@uw.edu" target="_blank">olzama@uw.edu</a>&gt;<br>
<b>Sent:</b> Tuesday, December 19, 2017 11:07 PM<br>
<b>To:</b> Woodley Packard<br>
<b>Cc:</b> <a href="mailto:developers@delph-in.net" target="_blank">developers@delph-in.net</a><br>
<b>Subject:</b> Re: [developers] post-reduction lexical gap</font>
<div> </div>
</div><div><div class="m_3599062933696227931m_-3720151325719300329m_-5267281124168586687h5">
<div>Hi Woodley,<br>
<br>
At a risk of asking a basic question: What does a lexical gap mean?<br>
<br>
Thank you,<br>
Olga<br>
<div class="m_3599062933696227931m_-3720151325719300329m_-5267281124168586687m_-3087037634506015501x_gmail_quote">
<div dir="ltr">On Tue, Dec 19, 2017 at 9:35 PM Woodley Packard &lt;<a href="mailto:sweaglesw@sweaglesw.org" target="_blank">sweaglesw@sweaglesw.org</a>&gt; wrote:<br>
</div>
<blockquote class="m_3599062933696227931m_-3720151325719300329m_-5267281124168586687m_-3087037634506015501x_gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div dir="auto">
<div>Hi all,</div>
<div><br>
</div>
<div>This means ACE encountered a lexical gap.  It should not be a rare occurrence, at least on non-toy data.</div>
<div><br>
</div>
<div>Best, Woodley<br>
<br>
<br>
</div>
</div>
<div dir="auto">
<div><br>
On Dec 19, 2017, at 1:21 PM, Stephan Oepen &lt;<a href="mailto:oe@ifi.uio.no" target="_blank">oe@ifi.uio.no</a>&gt; wrote:<br>
<br>
</div>
<blockquote type="cite">
<div>
<div>
<div dir="auto">hi olga,</div>
<div dir="auto"><br>
</div>
<div dir="auto">this looks like an error message from the parsing client, not [incr tsdb()] proper.  i would guess maybe ACE?  if so, woodley will likely be able to shed more light on this question :-).</div>
<div dir="auto"><br>
</div>
<div dir="auto">best, oe</div>
<div dir="auto"><br>
</div>
<br>
<div class="m_3599062933696227931m_-3720151325719300329m_-5267281124168586687m_-3087037634506015501x_gmail_quote">
<div>On Tue, 19 Dec 2017 at 19:19 Olga Zamaraeva &lt;<a href="mailto:olzama@uw.edu" target="_blank">olzama@uw.edu</a>&gt; wrote:<br>
</div>
<blockquote class="m_3599062933696227931m_-3720151325719300329m_-5267281124168586687m_-3087037634506015501x_gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div>Dear developers,<br>
<br>
What does the message &quot;post-reduction lexical gap&quot; mean in the context of adding an [ incr tsdb() ] profile for regression testing to the Grammar Matrix?<br>
<br>
Here&#39;s the type of line that we sometimes see in home/gold/mytest/parse<br>
<br>
10@1@10@-1@0@-1@0@0@-1@0@0@-1@0@-1@-1@-1@-1@-1@-1@-1@-1@-1@-1@0@0@0@0@0@-1@-1@-1@-1@34728@-1@-1@-1@23-6-2013 14:28:24@post-reduction lexical gap@<br>
<br>
This seems problematic but we don&#39;t really know what it is at this point.<br>
<br>
<div>(Here&#39;s a related topic on Discourse: <a href="https://delphinqa.ling.washington.edu/t/no-parses-when-adding-regression-tests/81/7" target="_blank">https://delphinqa.ling.washington.edu/t/no-parses-when-adding-regression-tests/81/7</a>)</div>
<div><br>
</div>
<div>Thank you,</div>
<div>Olga</div>
</div>
</blockquote>
</div>
</div>
</div>
</blockquote>
</div>
</blockquote>
</div>
</div>
</div></div></div>
</div>
</div>

</blockquote></div><br></div>
</div></blockquote></div></div></div></blockquote></div><br><br clear="all"><div><br></div></div></div><span>-- <br><div class="m_3599062933696227931m_-3720151325719300329gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr">Emily M. Bender<br>Professor, <span style="font-size:12.8px">Department of Linguistics</span></div><div dir="ltr">Check out CLMS on facebook! <a href="http://www.facebook.com/uwclma" target="_blank">http://www.facebook.com/uwclma</a><br></div></div></div></div></div></div>
</span></div>
</blockquote></div><br></div>
</blockquote></div></div></div>