<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!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>54.2. Reporting Errors Within the Server</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 Vsnapshot" /><link rel="prev" href="source-format.html" title="54.1. Formatting" /><link rel="next" href="error-style-guide.html" title="54.3. Error Message Style Guide" /></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">54.2. Reporting Errors Within the Server</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="source-format.html" title="54.1. Formatting">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="source.html" title="Chapter 54. PostgreSQL Coding Conventions">Up</a></td><th width="60%" align="center">Chapter 54. PostgreSQL Coding Conventions</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 11.5 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="error-style-guide.html" title="54.3. Error Message Style Guide">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="ERROR-MESSAGE-REPORTING"><div class="titlepage"><div><div><h2 class="title" style="clear: both">54.2. Reporting Errors Within the Server</h2></div></div></div><a id="id-1.10.6.3.2" class="indexterm"></a><a id="id-1.10.6.3.3" class="indexterm"></a><p>
Error, warning, and log messages generated within the server code
should be created using <code class="function">ereport</code>, or its older cousin
<code class="function">elog</code>. The use of this function is complex enough to
require some explanation.
</p><p>
There are two required elements for every message: a severity level
(ranging from <code class="literal">DEBUG</code> to <code class="literal">PANIC</code>) and a primary
message text. In addition there are optional elements, the most
common of which is an error identifier code that follows the SQL spec's
SQLSTATE conventions.
<code class="function">ereport</code> itself is just a shell function, that exists
mainly for the syntactic convenience of making message generation
look like a function call in the C source code. The only parameter
accepted directly by <code class="function">ereport</code> is the severity level.
The primary message text and any optional message elements are
generated by calling auxiliary functions, such as <code class="function">errmsg</code>,
within the <code class="function">ereport</code> call.
</p><p>
A typical call to <code class="function">ereport</code> might look like this:
</p><pre class="programlisting">
ereport(ERROR,
(errcode(ERRCODE_DIVISION_BY_ZERO),
errmsg("division by zero")));
</pre><p>
This specifies error severity level <code class="literal">ERROR</code> (a run-of-the-mill
error). The <code class="function">errcode</code> call specifies the SQLSTATE error code
using a macro defined in <code class="filename">src/include/utils/errcodes.h</code>. The
<code class="function">errmsg</code> call provides the primary message text. Notice the
extra set of parentheses surrounding the auxiliary function calls —
these are annoying but syntactically necessary.
</p><p>
Here is a more complex example:
</p><pre class="programlisting">
ereport(ERROR,
(errcode(ERRCODE_AMBIGUOUS_FUNCTION),
errmsg("function %s is not unique",
func_signature_string(funcname, nargs,
NIL, actual_arg_types)),
errhint("Unable to choose a best candidate function. "
"You might need to add explicit typecasts.")));
</pre><p>
This illustrates the use of format codes to embed run-time values into
a message text. Also, an optional <span class="quote">“<span class="quote">hint</span>”</span> message is provided.
</p><p>
If the severity level is <code class="literal">ERROR</code> or higher,
<code class="function">ereport</code> aborts the execution of the user-defined
function and does not return to the caller. If the severity level is
lower than <code class="literal">ERROR</code>, <code class="function">ereport</code> returns normally.
</p><p>
The available auxiliary routines for <code class="function">ereport</code> are:
</p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
<code class="function">errcode(sqlerrcode)</code> specifies the SQLSTATE error identifier
code for the condition. If this routine is not called, the error
identifier defaults to
<code class="literal">ERRCODE_INTERNAL_ERROR</code> when the error severity level is
<code class="literal">ERROR</code> or higher, <code class="literal">ERRCODE_WARNING</code> when the
error level is <code class="literal">WARNING</code>, otherwise (for <code class="literal">NOTICE</code>
and below) <code class="literal">ERRCODE_SUCCESSFUL_COMPLETION</code>.
While these defaults are often convenient, always think whether they
are appropriate before omitting the <code class="function">errcode()</code> call.
</p></li><li class="listitem"><p>
<code class="function">errmsg(const char *msg, ...)</code> specifies the primary error
message text, and possibly run-time values to insert into it. Insertions
are specified by <code class="function">sprintf</code>-style format codes. In addition to
the standard format codes accepted by <code class="function">sprintf</code>, the format
code <code class="literal">%m</code> can be used to insert the error message returned
by <code class="function">strerror</code> for the current value of <code class="literal">errno</code>.
<a href="#ftn.id-1.10.6.3.9.2.2.1.7" class="footnote"><sup class="footnote" id="id-1.10.6.3.9.2.2.1.7">[13]</sup></a>
<code class="literal">%m</code> does not require any
corresponding entry in the parameter list for <code class="function">errmsg</code>.
Note that the message string will be run through <code class="function">gettext</code>
for possible localization before format codes are processed.
</p></li><li class="listitem"><p>
<code class="function">errmsg_internal(const char *msg, ...)</code> is the same as
<code class="function">errmsg</code>, except that the message string will not be
translated nor included in the internationalization message dictionary.
This should be used for <span class="quote">“<span class="quote">cannot happen</span>”</span> cases that are probably
not worth expending translation effort on.
</p></li><li class="listitem"><p>
<code class="function">errmsg_plural(const char *fmt_singular, const char *fmt_plural,
unsigned long n, ...)</code> is like <code class="function">errmsg</code>, but with
support for various plural forms of the message.
<em class="replaceable"><code>fmt_singular</code></em> is the English singular format,
<em class="replaceable"><code>fmt_plural</code></em> is the English plural format,
<em class="replaceable"><code>n</code></em> is the integer value that determines which plural
form is needed, and the remaining arguments are formatted according
to the selected format string. For more information see
<a class="xref" href="nls-programmer.html#NLS-GUIDELINES" title="55.2.2. Message-writing Guidelines">Section 55.2.2</a>.
</p></li><li class="listitem"><p>
<code class="function">errdetail(const char *msg, ...)</code> supplies an optional
<span class="quote">“<span class="quote">detail</span>”</span> message; this is to be used when there is additional
information that seems inappropriate to put in the primary message.
The message string is processed in just the same way as for
<code class="function">errmsg</code>.
</p></li><li class="listitem"><p>
<code class="function">errdetail_internal(const char *msg, ...)</code> is the same
as <code class="function">errdetail</code>, except that the message string will not be
translated nor included in the internationalization message dictionary.
This should be used for detail messages that are not worth expending
translation effort on, for instance because they are too technical to be
useful to most users.
</p></li><li class="listitem"><p>
<code class="function">errdetail_plural(const char *fmt_singular, const char *fmt_plural,
unsigned long n, ...)</code> is like <code class="function">errdetail</code>, but with
support for various plural forms of the message.
For more information see <a class="xref" href="nls-programmer.html#NLS-GUIDELINES" title="55.2.2. Message-writing Guidelines">Section 55.2.2</a>.
</p></li><li class="listitem"><p>
<code class="function">errdetail_log(const char *msg, ...)</code> is the same as
<code class="function">errdetail</code> except that this string goes only to the server
log, never to the client. If both <code class="function">errdetail</code> (or one of
its equivalents above) and
<code class="function">errdetail_log</code> are used then one string goes to the client
and the other to the log. This is useful for error details that are
too security-sensitive or too bulky to include in the report
sent to the client.
</p></li><li class="listitem"><p>
<code class="function">errdetail_log_plural(const char *fmt_singular, const char
*fmt_plural, unsigned long n, ...)</code> is like
<code class="function">errdetail_log</code>, but with support for various plural forms of
the message.
For more information see <a class="xref" href="nls-programmer.html#NLS-GUIDELINES" title="55.2.2. Message-writing Guidelines">Section 55.2.2</a>.
</p></li><li class="listitem"><p>
<code class="function">errhint(const char *msg, ...)</code> supplies an optional
<span class="quote">“<span class="quote">hint</span>”</span> message; this is to be used when offering suggestions
about how to fix the problem, as opposed to factual details about
what went wrong.
The message string is processed in just the same way as for
<code class="function">errmsg</code>.
</p></li><li class="listitem"><p>
<code class="function">errcontext(const char *msg, ...)</code> is not normally called
directly from an <code class="function">ereport</code> message site; rather it is used
in <code class="literal">error_context_stack</code> callback functions to provide
information about the context in which an error occurred, such as the
current location in a PL function.
The message string is processed in just the same way as for
<code class="function">errmsg</code>. Unlike the other auxiliary functions, this can
be called more than once per <code class="function">ereport</code> call; the successive
strings thus supplied are concatenated with separating newlines.
</p></li><li class="listitem"><p>
<code class="function">errposition(int cursorpos)</code> specifies the textual location
of an error within a query string. Currently it is only useful for
errors detected in the lexical and syntactic analysis phases of
query processing.
</p></li><li class="listitem"><p>
<code class="function">errtable(Relation rel)</code> specifies a relation whose
name and schema name should be included as auxiliary fields in the error
report.
</p></li><li class="listitem"><p>
<code class="function">errtablecol(Relation rel, int attnum)</code> specifies
a column whose name, table name, and schema name should be included as
auxiliary fields in the error report.
</p></li><li class="listitem"><p>
<code class="function">errtableconstraint(Relation rel, const char *conname)</code>
specifies a table constraint whose name, table name, and schema name
should be included as auxiliary fields in the error report. Indexes
should be considered to be constraints for this purpose, whether or
not they have an associated <code class="structname">pg_constraint</code> entry. Be
careful to pass the underlying heap relation, not the index itself, as
<code class="literal">rel</code>.
</p></li><li class="listitem"><p>
<code class="function">errdatatype(Oid datatypeOid)</code> specifies a data
type whose name and schema name should be included as auxiliary fields
in the error report.
</p></li><li class="listitem"><p>
<code class="function">errdomainconstraint(Oid datatypeOid, const char *conname)</code>
specifies a domain constraint whose name, domain name, and schema name
should be included as auxiliary fields in the error report.
</p></li><li class="listitem"><p>
<code class="function">errcode_for_file_access()</code> is a convenience function that
selects an appropriate SQLSTATE error identifier for a failure in a
file-access-related system call. It uses the saved
<code class="literal">errno</code> to determine which error code to generate.
Usually this should be used in combination with <code class="literal">%m</code> in the
primary error message text.
</p></li><li class="listitem"><p>
<code class="function">errcode_for_socket_access()</code> is a convenience function that
selects an appropriate SQLSTATE error identifier for a failure in a
socket-related system call.
</p></li><li class="listitem"><p>
<code class="function">errhidestmt(bool hide_stmt)</code> can be called to specify
suppression of the <code class="literal">STATEMENT:</code> portion of a message in the
postmaster log. Generally this is appropriate if the message text
includes the current statement already.
</p></li><li class="listitem"><p>
<code class="function">errhidecontext(bool hide_ctx)</code> can be called to
specify suppression of the <code class="literal">CONTEXT:</code> portion of a message in
the postmaster log. This should only be used for verbose debugging
messages where the repeated inclusion of context would bloat the log
volume too much.
</p></li></ul></div><p>
</p><div class="note"><h3 class="title">Note</h3><p>
At most one of the functions <code class="function">errtable</code>,
<code class="function">errtablecol</code>, <code class="function">errtableconstraint</code>,
<code class="function">errdatatype</code>, or <code class="function">errdomainconstraint</code> should
be used in an <code class="function">ereport</code> call. These functions exist to
allow applications to extract the name of a database object associated
with the error condition without having to examine the
potentially-localized error message text.
These functions should be used in error reports for which it's likely
that applications would wish to have automatic error handling. As of
<span class="productname">PostgreSQL</span> 9.3, complete coverage exists only for
errors in SQLSTATE class 23 (integrity constraint violation), but this
is likely to be expanded in future.
</p></div><p>
There is an older function <code class="function">elog</code> that is still heavily used.
An <code class="function">elog</code> call:
</p><pre class="programlisting">
elog(level, "format string", ...);
</pre><p>
is exactly equivalent to:
</p><pre class="programlisting">
ereport(level, (errmsg_internal("format string", ...)));
</pre><p>
Notice that the SQLSTATE error code is always defaulted, and the message
string is not subject to translation.
Therefore, <code class="function">elog</code> should be used only for internal errors and
low-level debug logging. Any message that is likely to be of interest to
ordinary users should go through <code class="function">ereport</code>. Nonetheless,
there are enough internal <span class="quote">“<span class="quote">cannot happen</span>”</span> error checks in the
system that <code class="function">elog</code> is still widely used; it is preferred for
those messages for its notational simplicity.
</p><p>
Advice about writing good error messages can be found in
<a class="xref" href="error-style-guide.html" title="54.3. Error Message Style Guide">Section 54.3</a>.
</p><div class="footnotes"><br /><hr style="width:100; text-align:left;margin-left: 0" /><div id="ftn.id-1.10.6.3.9.2.2.1.7" class="footnote"><p><a href="#id-1.10.6.3.9.2.2.1.7" class="para"><sup class="para">[13] </sup></a>
That is, the value that was current when the <code class="function">ereport</code> call
was reached; changes of <code class="literal">errno</code> within the auxiliary reporting
routines will not affect it. That would not be true if you were to
write <code class="literal">strerror(errno)</code> explicitly in <code class="function">errmsg</code>'s
parameter list; accordingly, do not do so.
</p></div></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="source-format.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="source.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="error-style-guide.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">54.1. Formatting </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> 54.3. Error Message Style Guide</td></tr></table></div></body></html>
