Notice: The Concordance will now provide links to files that are resident at http://www.ucle.org.  This includes files in the "UCLE corner" and in the Audio Archive.  Perlfect Search will not provide links to those files.

AUE Web-Site Concordance Index

Help

To search for a word (hereinafter called the 'target word'), find an entry in the above index that is the highest in alphabetical order that is less than or equal to the target word.  Click on that entry.  This will bring up a concordance file that will contain the target word.  Search for the target word using your browser's search function.  (For example, in Internet Explorer or Netscape use control-F.)  If you prepend a colon to the target word, the search will skip all occurrences except the one that is a main entry.  The main entry for a target word in the concordance file will be followed by a list of hyperlinks, one for each search area (See footnote) that has at least one occurrence of the target word.  A number following each hyperlink indicates the number of occurrences of the target word in the search area represented by the hyperlink.  

For example, to search for the word 'machine', note the adjacent entries 'm' and 'mak' in the index.  The word 'machine' is alphabetically equal to or greater than 'm' and less than 'mak', so you should click on 'm'.

Each hyperlink in the Concordance is preceded by an italicized tag that stands for the program or the type of program that the link pertains to.  The following table shows some of the tags and the program or program type they correspond to.   Other tags should be self explanatory.

Added comment, July 2001:

With recent enhancesments to the Concordance feature, probably the most important distinction between the Concordance and Perlfect Search has to do with the ways John Lawler's AUE FAQ and Michael Quinion's World Wide Words are handled.

Concordance hits are now provided for the internal text of material at John Lawler's site. For example, for the string "hopefully", one Concordance hit takes you directly to John's site and the file "whom.html" in which "hopefully" occurs.

Hits pertaining to Michael Quinion's World Wide Words are based on index entries only, not on internal text. For example, for the word "hops", a Concordance hit reads "wwwords Mad as hops: 1", and that hit takes you directly to Michael Quinion's site and the article entitled "Mad as hops".

With respect to John's and Michael's sites, Perlfect Search works only on the echoed indexes of those sites. For example, a Perlfect search on "hops" takes you to the echoed index for World Wide Words, which you then have to search for "hops". That yields the entry "Mad as hops", and clicking on that entry takes you to Michael's site and the article with that title.

What all of this amounts to is that the Concordance will provide hits at John Lawler's site that Perlfect Search will not find, and the Concordance will provide more convenient access to Michael Quinion's site than will Perlfect Search.

Some people may find Perlfect preferable because it has somewhat the feel of search engines like AltaVista.  It also allows phrases to be searched for as well as individual words.

Other people may find the Concordance preferable because it takes advantage of internal anchors in some files so that it goes more directly to the point where the search target is to be found.

The Concordance doesn't provide for searching on multiword phrases, but it usually happens that if you search for a word in the phrase, one of the hits has the phrase spelled out in its hyperlink.  To see what I mean, use the Concordance to look for 'rule of thumb' by searching on 'thumb'; or 'cut the mustard', by searching on 'mustard'.

A difference that may be significant to people with slower systems is that each time you select a segment of the Concordance, a file is temporarily downloaded to your site.  An effort has been made to keep the sizes of the segments small enough so that this download time would not be prohibitive.  Perlfect Search uses a single large index, but when it's used it's not downloaded from the remote Web site, so download time is not a problem.

In the allocation of the overall Concordance to the individual concordance files, the rule used was that no concordance file should exceed 50 kilobytes in size.  This is the reason for some letters of the alphabet having more partitions than others.  It also explains the uneven spacing of the partitions.  An exception to the 50-kilobyte rule is the 'Non-alpha' file, which contains entries that don't start with an alphabetic character.

The Perl script that generates the Concordance was written by Markus Laker.