In this example, only one entry has been indexed in the main glossary. You can find out more information by clicking on the Details link, which will open the window shown in .

Since only one entry has been used, there's only one row. The first column lists the entry's label, the second column lists the entry's sort field and the third column shows the number of times that entry was indexed in the document. If you have a long list of entries, you can use the search box to find an entry according to its label. (The sort column isn't searched.) Regular expressions are permitted.

The tab () provides information, warnings and suggestions. In this example, there are no errors detected, so it just provides suggestions and some links on how to incorporate into your document build process. There are also two buttons provided to test the and scripts. In the first case, the action will also test if Perl is installed.

If you have defined an entry in your document, but it's not listed in the details window for the relevant glossary, then it hasn't been indexed in your document. Remember that the commands described in section9 (Using Glossary Terms Without Links) of the glossaries manual don't index the terms. These essentially are all the commands in the form glsentryfield or glossentrynamefield, such as , , or , and their case-changing variants. Also and .

If you're using the glossaries-extra package, remember that the noindex option will suppress indexing.

Now let's consider the following document (called, say, missing-sort.tex):

\documentclass{article}

\usepackage[utf8]{inputenc}
\usepackage[xindy]{glossaries}

\makeglossaries

\newglossaryentry{S}{name={\S},
 description={section symbol}}

\newglossaryentry{alpha}{name={\ensuremath{\alpha}},
 description={alpha}}

\newglossaryentry{beta}{name={$\beta$},text={\beta},
 description={beta}}

\begin{document}
Test: \gls{S}, $\gls{alpha}$, $\gls{beta}$.

\printglossaries

\end{document}
   

As before, run as usual on this document. Since the xindy package option has been used, this will create a .xdy file instead of a .ist file and the .glo file is now in 's format. There are, however, problems with this document. The glossaries manual advises using the sort key for entries that contain special characters or commands in the entry's name. This document hasn't followed that advice, and xindy will complain. The S entry just causes a warning:

Would replace complete index key with empty string, ignoring

and the S entry is ignored. The alpha and beta entries cause an error:

index 0 should be less than the length of the string

Again the entries are ignored, but the message is fairly cryptic. If we load the auxiliary file (missing-sort.aux) into , these problems are detected, and the following error message is displayed:

Xindy has ignored one or more entries with empty sort strings. Xindy failed with exit code 1.

Once this error message has been dismissed, the

tab should automatically be selected (see ). This identifies the problem entries and recommends a solution, in this case, add the sort key to the entry definition. The actual warning and error message reported by are shown at the end. (You can adjust the font used by these messages if you like, see .)

In the

panel, the Details link can again be used to view the list of indexed entries, but now the problematic entries are shown in red (see ).

actually performs more problem-checking than as it also tries to parse the log file for certain messages (but only in mode). This is illustrated in the next example, which won't generate any error messages from .

Spot what's wrong with the following document:

\documentclass{article}

\usepackage{glossaries}

\makeglossaries

\newglossaryentry{sample}{name={sample},
 description={an example}}

\begin{document}
A \gls{sample} document.

\printglossary[type=acronym]

\end{document}

If you use the normal method of , , , you won't get any error messages, but the glossary won't be displayed. Why not? If we switch from

\usepackage{glossaries}

to

\usepackage{glossaries-extra}

then we do finally get an error:

! Package glossaries-extra Error: Glossary type acronym doesn't exist.

The glossaries-extra package is stricter than the base glossaries package. The problem here is that I've used type=acronym, but there's no glossary with that label. (I haven't used the acronym option.) If you're not using the extension package, this is harder to pick up, but will notify you of the problem. This example document will trigger the error

No glossary acronym.

and the diagnostics panel will show the message:

It looks as though you might have done something like [type=acronym], but there's no acronym glossary.

If you switch to the

panel, the labels for the glossaries defined in the document are listed next to so you can check the indicated type against it. In this example, the list has only the single label main.

Remember that you not only have to define your entries, but you also have to index them if you want them to appear in the glossary. The glossaries package provides many commands that index entries, the most commonly used one being , which displays the text associated with the entry, indexes the entry, marks it as having been used and (if the hyperref package has been loaded) also creates a link to the definition in the glossary. Other commands provide variations, such as displaying different text or not changing the first use flag. In particular, the command only indexes the entry without displaying any text.

In the sample document below, I've defined an entry but it hasn't been indexed anywhere in the document.

\documentclass{article}

\usepackage{glossaries}

\makeglossaries

\newglossaryentry{sample}{name={sample},
 description={an example}}

\begin{document}
A sample document.

\printglossaries

\end{document}

My first step, as usual, is to run on this document, which will create the .aux file. Now if I try loading this file into I get the error message:

No entries were found for glossary main.

The diagnostics panel shows the following message:

There were no entries listed for the main glossary. Remember that you must index entries for them to appear in the glossary using the commands provided by the glossaries package. Entries that have been defined but not indexed won't be listed. If you don't want to use this glossary, add the nomain package option to your document. Check the following:

• Have you used commands like or in the document? (If you haven't, you need to add them.)
• If you have used commands like or in the preamble, have you remembered to put them after
• If you have at least version 4.24 of the glossaries package, have you used the debug option? (That might provide some more information for me to analyse.)

(The sentence referencing nomain only appears if there are no entries for the main glossary, but not for any other glossaries.)

Remember that if you use , you don't need or .

Consider the following document:

\documentclass{article}

\usepackage{glossaries}

\makenoidxglossaries

\newglossaryentry{sample}{name={sample},
 description={an example}}

\begin{document}
A \gls{sample} document.

\printnoidxglossaries

\end{document}

If I load the .aux file for this document into , I get the following message in the diagnostics panel:

It seems you've used , which means you don't need xindy or makeindex, you just need a second LaTeX run to get the glossary up to date.

Note that can still provide some limited diagnostics even when has been used. To illustrate this, if we modify the above sample document slightly, introducing an error:

\documentclass{article}

\usepackage{glossaries}

\makenoidxglossaries

\newglossaryentry{sample}{name={sample},
 description={an example}}

\begin{document}
A \gls{sample} document.

\printnoidxglossary[type=acronym]

\end{document}

This provides some additional information in the diagnostics panel:

Package glossaries Warning: Empty glossary for [type=acronym] Rerun may be required (or you may have forgotten to use commands like ) on input line 13.

It looks as though you might have done something like [type=acronym], but there's no acronym glossary.

So picks up the error.

Note that also looks for warnings from the glossaries package, so if you are encountering any problems, make sure you haven't suppressed the warnings with the nowarn package option.

In this example I've omitted from the document:

\documentclass{article}

\usepackage{glossaries}

\makeglossaries

\newglossaryentry{sample}{name={sample},
 description={an example}}

\begin{document}
A \gls{sample} document.

\end{document}

This doesn't cause any problems for as all the associated files have been created correctly. The document simply doesn't load the file generated by as there's no (or ). However the glossaries package does generate a warning, and this warning is picked up by and displayed in the diagnostics panel:

Package glossaries Warning: No or found. (Remove if you don't want any glossaries.) This document will not have a glossary.

If you suppress these warnings then can't help.

Sometimes things can go so badly wrong that doesn't even generate an auxiliary file. In this case you can load the .log file instead. (You'll need to change the file selector filter to show all files.) will parse the log file to see if it recognises any of the error messages. Some error messages can be quite crytic so there's no guarantee that will be able to help, but it might detect something useful. Note that this option is only available in mode.

Consider the following document:

\batchmode
\documentclass{beamer}

\usepackage{glossaries}

\makeglossaries

\newglossaryentry{sample}{name=sample,first={\textit{sample}},
  description={an example}}

\begin{document}

\begin{frame}
\gls{sample}
\end{frame}

\begin{frame}
\printglossary
\end{frame}

\end{document}

This document goes badly wrong. The first error message is:

! Undefined control sequence.
\in@ #1#2->\begingroup \def \in@@

If I load the log file into , the diagnostic panel displays the following:

Since the aux file doesn't exist, there's not much I can do to help, but I'll parse the log file in case there are any clues there.

It's possible that there's an expansion issue involving a fragile command. Things to check for:

• Have you used a class like beamer that doesn't make common formatting commands like robust?
• Have you tried using in front of fragile commands contained within your entry definitions?
• Have you tried switching off the expansion using commands like ? (See section 4.6 Expansion in the glossaries user manual.)

The problem here is that a fragile command has been used in the entry definition. The problematic command in this example is , which is normally robust, but it happens to be fragile with the beamer class. The solution is to either protect the problematic command with or use before you define the entries. For example:

\documentclass{beamer}

\usepackage{glossaries}

\makeglossaries

\glsnoexpandfields

\newglossaryentry{sample}{name=sample,first={\textit{sample}},
  description={an example}}

\begin{document}

\begin{frame}
\gls{sample}
\end{frame}

\begin{frame}
\printglossary
\end{frame}

\end{document}

The main panel shows the character encoding that believes is being used by the indexing application and, if detected, the document input encoding. Since only supports characters in the range 1 to 255, assumes ISO-8859-1 (Latin-1) for and will add an advisory note if the document class uses a different encoding.

Consider the following example:

\documentclass{article}

\usepackage[T1]{fontenc}
\usepackage[utf8]{inputenc}

\usepackage[style=indexgroup]{glossaries}

\makeglossaries

\newglossaryentry{elite}{name={élite},description={select group}}

\begin{document}
\gls{elite}

\printglossary
\end{document}

This document is saved as UTF-8 and has a term where the sort value starts with an extended character. This doesn't work as treats the UTF-8 letter as two separate characters form from the two octets. This not only affects the sorting but also causes a problem for the indexgroup style.

The first message in the diagnostics panel (see ) is only picked up after you rerun and reload the file in , as it's only after the glossary file has been created by that the call fails. The failure is caused by the first octet appearing in the argument of . This causes two problems: the argument of this command is a label so special or active characters will break it, and the inputenc package makes the first octet active, requiring the second octet as the argument. The message reads:

There seems to be a problem with the letter group label Ã. The label is used to construct a command name, so it can't contain any special characters. (This includes extended characters if you're using inputenc.sty.)

l.3 \glsgroupheading{Ã}

You may want to consider using xindy with a LaTeX engine that has native Unicode support (XeLaTeX or LuaLaTeX) or use bib2gls instead.

This message doesn't show up when you first attempt to create the glossary files with . However, there's advisory message than points to a problem:

The indexer encoding ISO-8859-1 doesn't seem to match the document encoding (utf8). This may not be a problem if you aren't using extended characters in the sort values.

In this case it is a problem. The two different encodings are also shown in . The indexer encoding is listed as ISO-8859-1, and the document encoding is listed as utf8.

The problem shows up more clearly in the Details window (see ), which uses the indexer's encoding and so displays the sort value as élite instead of élite.

The application settings can be adjusted through the

menu. This has menu items for increasing or decreasing the font size ( or ), setting the dry run mode () or open the dialog window (). Note that the dry run mode is the only setting that isn't remembered the next time you run .

The

dialog has four tabs: , , and .

The

() tab allows you to select the directory to use on start up. This is the directory the file chooser will be set to initially.

will try to determine if you have defined any entries within the environment. Although the glossaries package allows document definitions, the manual encourages defining entries in the preamble, see section4.10 of the glossaries user manual (Drawbacks With Defining Entries in the Document Environment). If you want to skip this check, deselect the

check box.

There's also a check to see if the glossaries package has complained about missing language modules. Not all languages are supported and, for those languages that are supported, the appropriate module must be installed in addition to installing the glossaries package. If the required language support is missing, the glossary files can still be built, you'll just have to manually change the fixed text for the title following the instructions in section1.3 (Multi-Lingual Support) of the glossaries user manual. If you want to skip this check, deselect the

check box.

The

tab lists the paths to and . will attempt to locate them on your system's path, but if they can't be detected, you'll need to specify the correct location. You can omit the location for an unrequired application.

will try to determine the language and input encoding from the .aux file to pass to , but you can override this if you want to. Make sure that the

check box is selected, and change the language and encoding as appropriate. Note that the batch mode will also use these settings, although they can only be adjusted in the mode.

The font used in the

and panels can be set in the tab. In addition to adjusting the font size through the or menu items, you can also set the required font size in this tab.

The can be set by selecting the required option in the

drop-down menu. Note that a restart is required as the Look and Feel is set on start up. The title bar appearance is governed by your usual operating system preference. The Look and Feel changes the way the window elements are displayed. shows the Metal Look and Feel. shows the Nimbus Look and Feel. shows the CDE/Motif Look and Feel. shows the GTK+ Look and Feel. You may not have all these options on your system or you may have additional options, depending on your Java installation.

is licensed under the terms of the GNU General Public License. depends on the following third party libraries whose jar files are in the lib directory: Java Help () and the Java Look and Feel Graphics Repository ().