BIND 10 trac2472, updated. 938ab8a39de2ae162b6e016bba718a2d6b874171 [2472] Merge branch 'master' into trac2472
BIND 10 source code commits
bind10-changes at lists.isc.org
Wed Nov 14 15:37:22 UTC 2012
The branch, trac2472 has been updated
via 938ab8a39de2ae162b6e016bba718a2d6b874171 (commit)
via e4fe372ea63d2115ca2e15866f0ef032fb05e0ca (commit)
via c4b0294b5bf4a9d32fb18ab62ca572f492788d72 (commit)
via 470aa3b8ecb855b0a9f6ad75b44c4f30bf35de8a (commit)
via be50a0c801d9203ca670e6cc9fa32e3461bc2d88 (commit)
via d654be880d463a899d2f6d68e0193c3ecac00500 (commit)
via 36378804a28850cb882bbc087c819fcabb5ada16 (commit)
via 8f4e6ef30b3078b8720b618efdacfb4fb186ad9f (commit)
via 3f1afa7c181c802ba2dab143cc9b5f72e8e1de69 (commit)
via 10d9112a1eefa19109036da9e4a39bc4ba836021 (commit)
via 9a11ef62ae36d9d891e87ba792fa249ae82f0736 (commit)
via 30b659a84d2d1beedc9adc53fbfd2ff71587dea5 (commit)
via 8872c15e883710475348463a87916d589c428031 (commit)
via 3d71b86e777c7280043a1191bbeb0bb115f88078 (commit)
via 15d6d71ed052c0dc5d6be39a8afe484014887b31 (commit)
via a3dde49a8210f65d1f37c9dee5638dc62ed92dd3 (commit)
via 6a2d1a5cbdeb2603982abbf148012e0879bba4ee (commit)
via 9fb295c549978ba217384643beb60b25b52ff35a (commit)
via a49d071d0040474aa3fb23cb0e79c92f51f54f71 (commit)
via 626a7de9ad0d9cc7324e024b41f55ef636ee7957 (commit)
via 56f0319afe0f63710358a2fd123e9865f1a13758 (commit)
via 38880a0a12721118c423dee2396ccfd165b103d0 (commit)
via aaf061f2c3f18d7e1cdc09c71ecb8820a107e64c (commit)
via 45088a0f850de1724eff8daff9fdc2303887ee13 (commit)
via 3446f7e74be156440c4e6b0333955b4c84e5443f (commit)
via 573c996957cb374223a921c0ec27a5218315b865 (commit)
via 3be2894bafa1ac7fdc9cbdfd81b9eb0b1aba16d6 (commit)
via 3ce2efa154f8eff8fd7029925ffd29116e2be8fa (commit)
via 36c280eef8a417854f7bfd2c7ad1356a8d52dcee (commit)
via 6e50ccbf961a340686ed9a6cc4011a7d9a867ccb (commit)
via 0c5e6acbe260c8095c3981526d3678e48f2faca9 (commit)
via 54ac6712c946aad84cd21330a08c3064e28ab91d (commit)
via 60977cf7528acb2bcc00067cb6a88a8136445453 (commit)
via cfc02b5aba598d12a89d51908beade5f3aaaf40f (commit)
via 5fd8dced290ab771f853ba6ae3127f16a2ac0689 (commit)
via c1799647fcc9e61a46b254da71ab0dcb24660885 (commit)
via 65dd63b2c9e72794b0029c5071e9fbf948321ccd (commit)
via 9478cb0fdc9896568928701bda3101b0a2fbcf16 (commit)
via d2103c888085b93e48731258fb7900c01a921734 (commit)
via 9fcdf4d524b54dc80a2b68eb800861f1727330c8 (commit)
via 182de87a7fd651f2b70e603b470ec71b8b3d48f0 (commit)
via d4902d224ce5d0ff1f7f289fd6ba32b0152dc5a3 (commit)
via a5767f8d4eaa3e94b157fc00211a0d0fc3a4a989 (commit)
via 65fee8d161688cc74df9f14b641b64bc5117cd1f (commit)
via 5044adfc3df675aad9dc287af4a3d799d40dfd80 (commit)
via b7939a2d63141a7920491f5f4c13c62eb13cf88c (commit)
via 65fffc8b2838d3d08d9aba54ad9b41099c9e32f7 (commit)
via 8baae428236f887b53988cf0c7b52c3275acd531 (commit)
via d6ed107c38459ad3c3f0b3f74475dda4707d7f23 (commit)
via 150c7dbba7294cf89dae06cafb3e07f407aa2062 (commit)
via 2645b4f705c2793e4a40bc4604a953da69ec2506 (commit)
from f57c0b95732d9e1c7f6d0cb8a91541d411eb66a3 (commit)
Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.
- Log -----------------------------------------------------------------
commit 938ab8a39de2ae162b6e016bba718a2d6b874171
Merge: f57c0b9 e4fe372
Author: Stephen Morris <stephen at isc.org>
Date: Wed Nov 14 14:44:23 2012 +0000
[2472] Merge branch 'master' into trac2472
-----------------------------------------------------------------------
Summary of changes:
ChangeLog | 5 +
configure.ac | 2 +-
doc/guide/bind10-guide.xml | 604 +++++++++++++++++++-
src/bin/auth/auth_messages.mes | 10 +
src/bin/auth/auth_srv.cc | 10 +-
src/bin/auth/datasrc_clients_mgr.h | 21 +-
.../auth/tests/datasrc_clients_builder_unittest.cc | 21 +-
src/bin/bindctl/moduleinfo.py | 3 +-
src/lib/dhcp/tests/Makefile.am | 3 -
src/lib/dns/Makefile.am | 1 +
src/lib/dns/master_lexer.cc | 172 +++++-
src/lib/dns/master_lexer.h | 29 +-
src/lib/dns/master_lexer_state.h | 144 +++++
src/lib/dns/tests/Makefile.am | 1 +
src/lib/dns/tests/master_lexer_state_unittest.cc | 256 +++++++++
tests/tools/perfdhcp/templates/Makefile.am | 2 -
16 files changed, 1245 insertions(+), 39 deletions(-)
create mode 100644 src/lib/dns/master_lexer_state.h
create mode 100644 src/lib/dns/tests/master_lexer_state_unittest.cc
-----------------------------------------------------------------------
diff --git a/ChangeLog b/ChangeLog
index 631c33c..087f5af 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+506. [doc] jelte
+ Added a chapter about the use of the bindctl command tool to
+ to the BIND 10 guide.
+ (Trac #2305, git c4b0294b5bf4a9d32fb18ab62ca572f492788d72)
+
505. [bug] jelte
Fixed a bug in b10-xfrin where a wrong call was made during the
final check of a TSIG-signed transfer, incorrectly rejecting the
diff --git a/configure.ac b/configure.ac
index f432499..95f95c6 100644
--- a/configure.ac
+++ b/configure.ac
@@ -767,7 +767,7 @@ if test "$MYSQL_CONFIG" != "" ; then
LIBS="$MYSQL_LIBS $LIBS"
AC_LINK_IFELSE(
- [AC_LANG_PROGRAM([#include <mysql/mysql.h>],
+ [AC_LANG_PROGRAM([#include <mysql.h>],
[MYSQL mysql_handle;
(void) mysql_init(&mysql_handle);
])],
diff --git a/doc/guide/bind10-guide.xml b/doc/guide/bind10-guide.xml
index 6065616..e0266e1 100644
--- a/doc/guide/bind10-guide.xml
+++ b/doc/guide/bind10-guide.xml
@@ -1275,6 +1275,14 @@ TODO
configuring BIND 10.
</para></note>
+ <note><para>
+ <command>bindctl</command> has an internal command history, as
+ well as tab-completion for most of the commands and arguments.
+ However, these are only enabled if the python readline module
+ is available on the system. If not, neither of these
+ features will be supported.
+ </para></note>
+
<para>
The <command>bindctl</command> tool provides an interactive
prompt for configuring, controlling, and querying the BIND 10
@@ -1284,22 +1292,590 @@ TODO
communicate to any other components directly.
</para>
-<!-- TODO: explain and show interface -->
+ <section id="bindctl_commandline_options">
+ <title>bindctl command-line options</title>
+ <variablelist>
+ <varlistentry>
+ <term>-a <replaceable><address></replaceable>, --address=<replaceable><address></replaceable></term>
+ <listitem>
+ <simpara>
+ IP address that BIND 10's <command>b10-cmdctl</command>
+ module is listening on. By default, this is 127.0.0.1.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>-c <replaceable><certificate file></replaceable>, --certificate-chain=<replaceable><certificate file></replaceable></term>
+ <listitem>
+ <simpara>
+ PEM-formatted server certificate file. When this option is
+ given, <command>bindctl</command> will verify the server
+ certificate using the given file as the root of the
+ certificate chain. If not specified, <command>bindctl
+ </command> does not validate the certificate.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>--csv-file-dir=<replaceable><csv file></replaceable></term>
+ <listitem>
+ <simpara>
+ <command>bindctl</command> stores the username and
+ password for logging in in a file called
+ <filename>default_user.csv</filename>;
+ this option specifies the directory where this file is
+ stored and read from. When not specified,
+ <filename>~/.bind10/</filename> is used.
+ <note>Currently, this file contains an unencrypted password.</note>
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>-h, --help</term>
+ <listitem>
+ <simpara>
+ Shows a short overview of the command-line options of
+ <command>bindctl</command>, and exits.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>--version</term>
+ <listitem>
+ <simpara>
+ Shows the version of <command>bindctl</command>, and exits.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>-p <replaceable><port number></replaceable>, --port=<replaceable><port number></replaceable></term>
+ <listitem>
+ <simpara>
+ Port number that BIND 10's <command>b10-cmdctl</command>
+ module is listening on. By default, this is port 8080.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ </variablelist>
+ </section>
- <para>
- Configuration changes are actually commands to
- <command>b10-cfgmgr</command>. So when <command>bindctl</command>
- sends a configuration, it is sent to <command>b10-cmdctl</command>
- (over a HTTPS connection); then <command>b10-cmdctl</command>
- sends the command (over a <command>b10-msgq</command> command
- channel) to <command>b10-cfgmgr</command> which then stores
- the details and relays (over a <command>b10-msgq</command> command
- channel) the configuration on to the specified module.
- </para>
+ <section id="bindctl_general_syntax">
+ <title>General syntax of bindctl commands</title>
+ The <command>bindctl</command> tool is an interactive
+ command-line tool, with dynamic commands depending on the
+ BIND 10 modules that are running. There are a number of
+ fixed commands that have no module and that are always
+ available.
- <para>
- </para>
+ The general syntax of a command is
+
+ <screen><userinput><module> <command> <replaceable>[argument(s)]</replaceable></userinput></screen>
+
+ For example, the Boss module has a 'shutdown' command to shut down
+ BIND 10, with an optional argument 'help':
+
+ <screen><userinput>> Boss shutdown help</userinput>
+Command shutdown (Shut down BIND 10)
+ help (Get help for command)
+This command has no parameters
+ </screen>
+ There are no mandatory arguments, only the optional 'help'.
+ </section>
+
+ <section id="bindctl_help">
+ <title>Bindctl help</title>
+ <command>help</command> is both a command and an option that is available to all other commands. When run as a command directly, it shows the available modules.
+ <screen>> <userinput>help</userinput>
+usage: <module name> <command name> [param1 = value1 [, param2 = value2]]
+Type Tab character to get the hint of module/command/parameters.
+Type "help(? h)" for help on bindctl.
+Type "<module_name> help" for help on the specific module.
+Type "<module_name> <command_name> help" for help on the specific command.
+
+Available module names:
+<emphasis>(list of modules)</emphasis>
+ </screen>
+
+ When 'help' is used as a command to a module, it shows the supported commands for the module; for example:
+ <screen>> <userinput>Boss help</userinput>
+Module Boss Master process
+Available commands:
+ help Get help for module.
+ shutdown Shut down BIND 10
+ ping Ping the boss process
+ show_processes
+ List the running BIND 10 processes
+ </screen>
+
+ And when added to a module command, it shows the description and parameters of that specific command; for example:
+ <screen>> <userinput>Auth loadzone help</userinput>
+Command loadzone ((Re)load a specified zone)
+ help (Get help for command)
+Parameters:
+ class (string, optional)
+ origin (string, mandatory)
+ </screen>
+
+ </section>
+ <section id="bindctl_command_arguments">
+ <title>Command arguments</title>
+ <simpara>
+ Commands can have arguments, which can be either optional or
+ mandatory. They can be specified by name
+ (e.g. <command><replaceable><command></replaceable> <replaceable><argument name>=<argument value></replaceable></command>), or positionally,
+ (e.g. <command><replaceable><command></replaceable> <replaceable><argument value 1></replaceable> <replaceable><argument value 2></replaceable></command>).
+ </simpara>
+ <simpara>
+ <command><replaceable><command></replaceable> <replaceable>help</replaceable></command>
+ shows the arguments a command supports and which of those are
+ mandatory, and in which order the arguments are expected if
+ positional arguments are used.
+ </simpara>
+ <simpara>
+ For example, the <command>loadzone</command> command of the Auth
+ module, as shown in the last example of the previous section, has
+ two arguments, one of which is optional. The positional arguments in
+ this case are class first and origin second; for example:
+ <screen>> <userinput>Auth loadzone IN example.com.</userinput></screen>
+ But since the class is optional (defaulting to IN), leaving it out
+ works as well:
+ <screen>> <userinput>Auth loadzone example.com.</userinput></screen>
+ </simpara>
+ <simpara>
+ The arguments can also be provided with their names, in which
+ case the order does not matter:
+ <screen>> <userinput>Auth loadzone origin="example.com." class="IN"</userinput></screen>
+ </simpara>
+ </section>
+
+ <section id="bindctl_module_commands">
+ <title>Module commands</title>
+ Each module has its own set of commands (if any), which will only be
+ available if the module is running. For instance, the
+ Auth module has a <command>loadzone</command> command.
+ The commands a module provides are documented in
+ this guide in the section of that module or in the module's
+ corresponding manual page.
+ </section>
+
+ <section>
+ <title>Configuration commands</title>
+ Configuration commands are used to view and change the configuration
+ of BIND 10 and its modules. Module configuration is only shown if
+ that module is running, but similar to commands, there are a number
+ of top-level configuration items that are always available (for
+ instance <varname>tsig_keys</varname> and
+ <varname>data_sources</varname>).
+
+ Configuration changes (set, unset, add and remove) are done locally
+ first, and have no immediate effect. The changes can be viewed with
+ <command>config diff</command>, and either reverted
+ (<command>config revert</command>), or committed
+ (<command>config commit</command>).
+ In the latter case, all local changes are submitted
+ to the configuration manager, which verifies them, and if they are
+ accepted, applied and saved in persistent storage.
+
+ When identifying items in configuration commands, the format is
+ <screen><userinput>Module/example/item</userinput></screen>
+ Sub-elements of names, lists and sets (see <xref linkend=
+ "bindctl_configuration_data_types"/>) are separated with the '/'
+ character, and list indices are identified with [<replaceable><index></replaceable>]; for example:
+
+ <screen><userinput>Module/example/list[2]/foo</userinput></screen>
+
+ <section id="bindctl_configuration_command_list">
+ <title>List of configuration commands</title>
+ The following configuration commands are available:
+ <variablelist>
+ <varlistentry>
+ <term>show [all] [item name]</term>
+ <listitem>
+ <simpara>
+ Shows the current configuration of the given item. If 'all'
+ is given, it will recurse through the entire set, and show
+ every nested value.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>show_json [item name]</term>
+ <listitem>
+ <simpara>
+ Shows the full configuration of the given item in JSON format.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>add <item name> [value]</term>
+ <listitem>
+ <simpara>
+ Add an entry to configuration list or a named set (see <xref
+ linkend="bindctl_configuration_data_types"/>).
+ When adding to a list, the command has one optional
+ argument, a value to add to the list. The value must
+ be in correct JSON and complete. When adding to a
+ named set, it has one mandatory parameter (the name to
+ add), and an optional parameter value, similar to when
+ adding to a list. In either case, when no value is
+ given, an entry will be constructed with default
+ values.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>remove</term>
+ <listitem>
+ <simpara>
+ Remove an item from a configuration list or a named set.
+ When removing an item for a list, either the index needs to
+ be specified, or the complete value of the element to remove
+ must be specified (in JSON format).
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>set <item name> <value></term>
+ <listitem>
+ <simpara>
+ Directly set the value of the given item to the given value.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>unset <item name></term>
+ <listitem>
+ <simpara>
+ Remove any user-specified value for the given item.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>diff</term>
+ <listitem>
+ <simpara>
+ Show all current local changes that have not been
+ committed yet.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>revert</term>
+ <listitem>
+ <simpara>
+ Revert all local changes without committing them.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>commit</term>
+ <listitem>
+ <simpara>
+ Send all local changes to the configuration manager, which
+ will validate them, and apply them if validation succeeds.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>go</term>
+ <listitem>
+ <simpara>
+ Go to a specific configuration part, similar to the 'cd'
+ command in a shell.
+ <note>There are a number of problems with the current
+ implementation of go within <command>bindctl</command>,
+ and we recommend not using it for general cases.</note>
+ </simpara>
+ </listitem>
+ </varlistentry>
+ </variablelist>
+ </section>
+
+ <section id="bindctl_configuration_data_types">
+ <title>Configuration data types</title>
+ Configuration data can be of different types, which can be modified
+ in ways that depend on the types. There are a few syntax
+ restrictions on these types, but only basic ones. Modules may impose
+ additional restrictions on the values of elements.
+ <variablelist>
+ <varlistentry>
+ <term>integer</term>
+ <listitem>
+ <simpara>
+ A basic integer; can be set directly with <command>config set</command>, to any integer value.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>real</term>
+ <listitem>
+ <simpara>
+ A basic floating point number; can be set directly with <command>config set</command>, to any floating point value.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>boolean</term>
+ <listitem>
+ <simpara>
+ A basic boolean value; can be set directly with <command>config set</command>, to either <command>true</command> or <command>false</command>.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>string</term>
+ <listitem>
+ <simpara>
+ A basic string value; can be set directly with <command>config set,</command> so any string. Double quotation marks are optional.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>null</term>
+ <listitem>
+ <simpara>
+ This is a special type representing 'no value at all'; usable in compound structures that have optional elements that are not set.
+ </simpara>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term>maps</term>
+ <listitem>
+ <simpara>
+ Maps are (pre-defined) compound collections of other
+ elements of any other type. They are not usually
+ modified directly, but their elements are. Every
+ top-level element for a module is a map containing
+ the configuration values for that map, which can
+ themselves be maps again. For instance, the Auth
+ module configuration is a map containing the
+ elements '<varname>listen_on</varname>' (list) and '<varname>tcp_recv_timeout</varname>'
+ (integer). When changing one of its values, they can
+ be modified directly with <command>config set
+ Auth/tcp_recv_timeout 3000</command>.
+ </simpara>
+ <simpara>
+ Some map entries are optional. If they are, and
+ currently have a value, the value can be unset by
+ using either <command>config unset
+ <replaceable><item name></replaceable>
+ </command> or <command>config set
+ <replaceable><item name></replaceable>
+ null</command>.
+
+ </simpara>
+ <simpara>
+ Maps <emphasis>can</emphasis> be modified as a whole,
+ but using the full JSON representation of
+ the entire map to set.
+
+ Since this involves a lot of text, this is usually
+ not recommended.
+ </simpara>
+ <simpara>
+ Another example is the Logging virtual module, which
+ is, like any module, a map, but it only contains one
+ element: a list of loggers. Normally, an
+ administrator would only modify that list (or its
+ elements) directly, but it is possible to set the
+ entire map in one command; for example:
+ <command> config set Logging { "loggers": [] } </command>
+ </simpara>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term>list</term>
+ <listitem>
+ <simpara>
+ A list is a compound list of other elements of the
+ same type. Elements can be added with <command>config
+ add <replaceable><list name> [value]</replaceable></command>, and removed with
+ <command>config remove <replaceable><list name> [value]</replaceable></command> or
+ <command>config remove <replaceable><list name></replaceable><replaceable><index></replaceable></command>.
+ The index is of the form <emphasis>square bracket, number,
+ square bracket</emphasis> (e.g.
+ <command>[0]</command>), and it immediately follows
+ the list name (there is no separator or space
+ between them). List indices start with 0 for the
+ first element.
+ </simpara>
+ <simpara>
+ For addition, if the value is omitted, an entry with
+ default values will be added. For removal, either
+ the index or the full value (in JSON format) needs
+ to be specified.
+ </simpara>
+ <simpara>
+ Lists can also be used with
+ <command>config set</command>,
+ but like maps, only by specifying the
+ entire list value in JSON format.
+ </simpara>
+ <simpara>
+ For example, this command shows the port number used for the second element of the list <varname>listen_on</varname> in the Auth module:
+ <command> config show Auth/listen_on[1]/port</command>
+ </simpara>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term>named set</term>
+ <listitem>
+ <simpara>
+ Named sets are similar to lists, in that they are
+ sets of elements of the same type, but they are not
+ indexed by numbers, but by strings.
+ </simpara>
+ <simpara>
+ Values can be added with
+ <command>config add <replaceable><item name> <string> [value]</replaceable></command>
+ where 'string' is the name of the element. If 'value'
+ is ommitted, default values will be used. Elements
+ can be removed with <command>config remove
+ <replaceable><item
+ name> <string></replaceable></command>
+ </simpara>
+ <simpara>
+ Elements in a named set can be addressed similarly
+ to maps.
+ </simpara>
+ <simpara>
+ For example, the <command>Boss/components</command>
+ elements is a named set;
+ adding, showing, and then removing an element
+ can be done with the following three commands (note
+ the '/'-character versus the space before
+ 'example_module'):
+ </simpara>
+ <simpara>
+ <command>config add Boss/components example_module</command>
+ </simpara>
+ <simpara>
+ <command>config show Boss/components/example_module</command>
+ </simpara>
+ <simpara>
+ <command>config remove Boss/components example_module</command>
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>any</term>
+ <listitem>
+ <simpara>
+ The 'any' type is a special type that can have any
+ form. Apart from that, it must consist of elements as
+ described in this chapter, there is no restriction
+ on which element types are used. This type is used
+ in places where different data formats could be
+ used. Element modification commands depend on the
+ actual type of the value. For instance, if the value
+ of an 'any' element is a list, <command>config add
+ </command> and <command>config remove</command> work
+ as for other lists.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ </variablelist>
+ </section>
+ </section>
+
+ <section>
+ <title>The execute command</title>
+ The <command>execute</command> command executes a set of commands,
+ either from a file
+ or from a pre-defined set. Currently, the only predefined set is
+ <command>init_authoritative_server</command>, which adds
+ <command>b10-auth</command>, <command>b10-xfrin</command>, and
+ <command>b10-xfrout</command> to the set of components to be
+ started by BIND 10. This
+ pre-defined set does not commit the changes, so these modules do not
+ show up for commands or configuration until the user enters
+ <command>config commit</command> after
+ <command>execute init_authoritative_server</command>. For example:
+
+ <screen>> <userinput>execute init_authoritative_server</userinput></screen>
+
+ <screen>> <userinput>execute file /tmp/example_commands</userinput></screen>
+
+ The optional argument <command>show</command> displays the exact set of
+ commands that would be executed; for example:
+
+ <screen>> <userinput>execute init_authoritative_server show</userinput>
+!echo adding Authoritative server component
+config add /Boss/components b10-auth
+config set /Boss/components/b10-auth/kind needed
+config set /Boss/components/b10-auth/special auth
+!echo adding Xfrin component
+config add /Boss/components b10-xfrin
+config set /Boss/components/b10-xfrin/address Xfrin
+config set /Boss/components/b10-xfrin/kind dispensable
+!echo adding Xfrout component
+config add /Boss/components b10-xfrout
+config set /Boss/components/b10-xfrout/address Xfrout
+config set /Boss/components/b10-xfrout/kind dispensable
+!echo adding Zone Manager component
+config add /Boss/components b10-zonemgr
+config set /Boss/components/b10-zonemgr/address Zonemgr
+config set /Boss/components/b10-zonemgr/kind dispensable
+!echo Components added. Please enter "config commit" to
+!echo finalize initial setup and run the components.
+ </screen>
+
+ The optional <command>show</command> argument may also be used when
+ executing a script from a file; for example:
+
+ <screen><userinput>> execute file /tmp/example_commands show</userinput></screen>
+
+ <section id="bindctl_execute_directives">
+ <title>Execute directives</title>
+ Within sets of commands to be run with the <command>execute</command>
+ command, a number of directives are supported:
+ <variablelist>
+ <varlistentry>
+ <term>!echo <replaceable><string></replaceable></term>
+ <listitem>
+ <simpara>
+ Prints the given string to <command>bindctl</command>'s
+ output.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>!verbose on</term>
+ <listitem>
+ <simpara>
+ Enables verbose mode; all following commands that are to
+ be executed are also printed.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ <varlistentry>
+ <term>!verbose off</term>
+ <listitem>
+ <simpara>
+ Disables verbose mode; following commands that are to
+ be executed are no longer printed.
+ </simpara>
+ </listitem>
+ </varlistentry>
+ </variablelist>
+ </section>
+
+ <section id="bindctl_execute_notes">
+ <title>Notes on execute scripts</title>
+ Within scripts, you can add or remove modules with the normal
+ configuration commands for <command>Boss/components</command>.
+ However, as module
+ configuration and commands do not show up until the module is
+ running, it is currently not possible to add a module and set
+ its configuration in one script. This will be addressed in the
+ future, but for now the only option is to add and configure
+ modules in separate commands and execute scripts.
+ </section>
+ </section>
</chapter>
<chapter id="common">
@@ -2990,7 +3566,7 @@ Dhcp6/subnet6 [] list (default)</screen>
enclosed in square brackets, even though only one range of addresses
is specified.</para>
<para>It is possible to define more than one pool in a
- subnet: continuing the previous example, further assume that
+ subnet: continuing the previous example, further assume that
2001:db8:1:0:5::/80 should be also be managed by the server. It could be written as
2001:db8:1:0:5:: to 2001:db8:1::5:ffff:ffff:ffff, but typing so many 'f's
is cumbersome. It can be expressed more simply as 2001:db8:1:0:5::/80. Both
diff --git a/src/bin/auth/auth_messages.mes b/src/bin/auth/auth_messages.mes
index 3780499..ac9ffb9 100644
--- a/src/bin/auth/auth_messages.mes
+++ b/src/bin/auth/auth_messages.mes
@@ -98,6 +98,16 @@ This debug message is issued when the separate thread for maintaining data
source clients successfully loaded the named zone of the named class as a
result of the 'loadzone' command.
+% AUTH_DATASRC_CLIENTS_BUILDER_LOAD_ZONE_NOCACHE skipped loading zone %1/%2 due to no in-memory cache
+This debug message is issued when the separate thread for maintaining data
+source clients received a command to reload a zone but skipped it because
+the specified zone is not loaded in-memory (but served from an underlying
+data source). This could happen if the loadzone command is manually issued
+by a user but the zone name is misspelled, but a more likely cause is
+that the command is sent from another BIND 10 module (such as xfrin or DDNS).
+In the latter case it can be simply ignored because there is no need
+for explicit reloading.
+
% AUTH_DATASRC_CLIENTS_BUILDER_RECONFIGURE_CONFIG_ERROR Error in data source configuration: %1
The thread for maintaining data source clients has received a command to
reconfigure, but the parameter data (the new configuration) contains an
diff --git a/src/bin/auth/auth_srv.cc b/src/bin/auth/auth_srv.cc
index dca8fd0..26a8489 100644
--- a/src/bin/auth/auth_srv.cc
+++ b/src/bin/auth/auth_srv.cc
@@ -651,9 +651,10 @@ AuthSrvImpl::processNormalQuery(const IOMessage& io_message, Message& message,
local_edns->setUDPSize(AuthSrvImpl::DEFAULT_LOCAL_UDPSIZE);
message.setEDNS(local_edns);
}
- // Get access to data source client list through the holder and keep the
- // holder until the processing and rendering is done to avoid inter-thread
- // race.
+
+ // Get access to data source client list through the holder and keep
+ // the holder until the processing and rendering is done to avoid
+ // race with any other thread(s) such as the background loader.
auth::DataSrcClientsMgr::Holder datasrc_holder(datasrc_clients_mgr_);
try {
@@ -688,6 +689,9 @@ AuthSrvImpl::processNormalQuery(const IOMessage& io_message, Message& message,
return (true);
// The message can contain some data from the locked resource. But outside
// this method, we touch only the RCode of it, so it should be safe.
+
+ // Lock on datasrc_clients_mgr_ acquired by datasrc_holder is
+ // released here upon its deletion.
}
bool
diff --git a/src/bin/auth/datasrc_clients_mgr.h b/src/bin/auth/datasrc_clients_mgr.h
index 38bf162..5bbdb99 100644
--- a/src/bin/auth/datasrc_clients_mgr.h
+++ b/src/bin/auth/datasrc_clients_mgr.h
@@ -581,6 +581,9 @@ DataSrcClientsBuilderBase<MutexType, CondVarType>::doLoadZone(
try {
boost::shared_ptr<datasrc::memory::ZoneWriter> zwriter =
getZoneWriter(*client_list, rrclass, origin);
+ if (!zwriter) {
+ return;
+ }
zwriter->load(); // this can take time but doesn't cause a race
{ // install() can cause a race and must be in a critical section
@@ -614,8 +617,14 @@ DataSrcClientsBuilderBase<MutexType, CondVarType>::getZoneWriter(
datasrc::ConfigurableClientList& client_list,
const dns::RRClass& rrclass, const dns::Name& origin)
{
- const datasrc::ConfigurableClientList::ZoneWriterPair writerpair =
- client_list.getCachedZoneWriter(origin);
+ // getCachedZoneWriter() could get access to an underlying data source
+ // that can cause a race condition with the main thread using that data
+ // source for lookup. So we need to protect the access here.
+ datasrc::ConfigurableClientList::ZoneWriterPair writerpair;
+ {
+ typename MutexType::Locker locker(*map_mutex_);
+ writerpair = client_list.getCachedZoneWriter(origin);
+ }
switch (writerpair.first) {
case datasrc::ConfigurableClientList::ZONE_SUCCESS:
@@ -626,8 +635,10 @@ DataSrcClientsBuilderBase<MutexType, CondVarType>::getZoneWriter(
<< "/" << rrclass << ": not found in any configured "
"data source.");
case datasrc::ConfigurableClientList::ZONE_NOT_CACHED:
- isc_throw(InternalCommandError, "failed to load zone " << origin
- << "/" << rrclass << ": not served from memory");
+ LOG_DEBUG(auth_logger, DBG_AUTH_OPS,
+ AUTH_DATASRC_CLIENTS_BUILDER_LOAD_ZONE_NOCACHE)
+ .arg(origin).arg(rrclass);
+ break; // return NULL below
case datasrc::ConfigurableClientList::CACHE_DISABLED:
// This is an internal error. Auth server must have the cache
// enabled.
@@ -636,8 +647,6 @@ DataSrcClientsBuilderBase<MutexType, CondVarType>::getZoneWriter(
"is somehow disabled");
}
- // all cases above should either return or throw, but some compilers
- // still need a return statement
return (boost::shared_ptr<datasrc::memory::ZoneWriter>());
}
} // namespace datasrc_clientmgr_internal
diff --git a/src/bin/auth/tests/datasrc_clients_builder_unittest.cc b/src/bin/auth/tests/datasrc_clients_builder_unittest.cc
index 585e7c3..838e032 100644
--- a/src/bin/auth/tests/datasrc_clients_builder_unittest.cc
+++ b/src/bin/auth/tests/datasrc_clients_builder_unittest.cc
@@ -308,8 +308,12 @@ TEST_F(DataSrcClientsBuilderTest, loadZone) {
"{\"class\": \"IN\","
" \"origin\": \"test1.example\"}"));
EXPECT_TRUE(builder.handleCommand(loadzone_cmd));
- EXPECT_EQ(1, map_mutex.lock_count); // we should have acquired the lock
- EXPECT_EQ(1, map_mutex.unlock_count); // and released it.
+
+ // loadZone involves two critical sections: one for getting the zone
+ // writer, and one for actually updating the zone data. So the lock/unlock
+ // count should be incremented by 2.
+ EXPECT_EQ(2, map_mutex.lock_count);
+ EXPECT_EQ(2, map_mutex.unlock_count);
newZoneChecks(clients_map, rrclass);
}
@@ -381,7 +385,10 @@ TEST_F(DataSrcClientsBuilderTest,
find(Name("example.org")).finder_->
find(Name("example.org"), RRType::SOA())->code);
- // attempt of reloading a zone but in-memory cache is disabled.
+ // attempt of reloading a zone but in-memory cache is disabled. In this
+ // case the command is simply ignored.
+ const size_t orig_lock_count = map_mutex.lock_count;
+ const size_t orig_unlock_count = map_mutex.unlock_count;
const ConstElementPtr config2(Element::fromJSON("{"
"\"IN\": [{"
" \"type\": \"sqlite3\","
@@ -390,11 +397,13 @@ TEST_F(DataSrcClientsBuilderTest,
" \"cache-zones\": [\"example.org\"]"
"}]}"));
clients_map = configureDataSource(config2);
- EXPECT_THROW(builder.handleCommand(
+ builder.handleCommand(
Command(LOADZONE, Element::fromJSON(
"{\"class\": \"IN\","
- " \"origin\": \"example.org\"}"))),
- TestDataSrcClientsBuilder::InternalCommandError);
+ " \"origin\": \"example.org\"}")));
+ // Only one mutex was needed because there was no actual reload.
+ EXPECT_EQ(orig_lock_count + 1, map_mutex.lock_count);
+ EXPECT_EQ(orig_unlock_count + 1, map_mutex.unlock_count);
// basically impossible case: in-memory cache is completely disabled.
// In this implementation of manager-builder, this should never happen,
diff --git a/src/bin/bindctl/moduleinfo.py b/src/bin/bindctl/moduleinfo.py
index 6c3a304..33114d8 100644
--- a/src/bin/bindctl/moduleinfo.py
+++ b/src/bin/bindctl/moduleinfo.py
@@ -221,7 +221,8 @@ class ModuleInfo:
def module_help(self):
"""Prints the help info for this module to stdout"""
- print("Module ", self, "\nAvailable commands:")
+ print("Module " + str(self))
+ print("Available commands:")
for k in self.commands.values():
n = k.get_name()
if len(n) >= CONST_BINDCTL_HELP_INDENT_WIDTH:
diff --git a/src/lib/dhcp/tests/Makefile.am b/src/lib/dhcp/tests/Makefile.am
index bf5e2e7..f038fa0 100644
--- a/src/lib/dhcp/tests/Makefile.am
+++ b/src/lib/dhcp/tests/Makefile.am
@@ -4,9 +4,6 @@ AM_CPPFLAGS = -I$(top_builddir)/src/lib -I$(top_srcdir)/src/lib
AM_CPPFLAGS += $(BOOST_INCLUDES)
AM_CPPFLAGS += -DTEST_DATA_BUILDDIR=\"$(abs_top_builddir)/src/lib/dhcp/tests\"
AM_CPPFLAGS += -DINSTALL_PROG=\"$(abs_top_srcdir)/install-sh\"
-# Temp
-AM_CPPFLAGS += -I/usr/include/mysql -DBIG_JOINS=1 -fno-strict-aliasing
-AM_CPPFLAGS += -DUNIV_LINUX
AM_CXXFLAGS = $(B10_CXXFLAGS)
diff --git a/src/lib/dns/Makefile.am b/src/lib/dns/Makefile.am
index e81ef76..14b74f7 100644
--- a/src/lib/dns/Makefile.am
+++ b/src/lib/dns/Makefile.am
@@ -97,6 +97,7 @@ libb10_dns___la_SOURCES += master_lexer_inputsource.h master_lexer_inputsource.c
libb10_dns___la_SOURCES += labelsequence.h labelsequence.cc
libb10_dns___la_SOURCES += masterload.h masterload.cc
libb10_dns___la_SOURCES += master_lexer.h master_lexer.cc
+libb10_dns___la_SOURCES += master_lexer_state.h
libb10_dns___la_SOURCES += message.h message.cc
libb10_dns___la_SOURCES += messagerenderer.h messagerenderer.cc
libb10_dns___la_SOURCES += name.h name.cc
diff --git a/src/lib/dns/master_lexer.cc b/src/lib/dns/master_lexer.cc
index c9c5528..0978a8f 100644
--- a/src/lib/dns/master_lexer.cc
+++ b/src/lib/dns/master_lexer.cc
@@ -16,6 +16,7 @@
#include <dns/master_lexer.h>
#include <dns/master_lexer_inputsource.h>
+#include <dns/master_lexer_state.h>
#include <boost/shared_ptr.hpp>
@@ -32,10 +33,34 @@ typedef boost::shared_ptr<master_lexer_internal::InputSource> InputSourcePtr;
using namespace master_lexer_internal;
struct MasterLexer::MasterLexerImpl {
- MasterLexerImpl() : token_(Token::NOT_STARTED) {}
+ MasterLexerImpl() : source_(NULL), token_(Token::NOT_STARTED),
+ paren_count_(0), last_was_eol_(false)
+ {}
+
+ // A helper method to skip possible comments toward the end of EOL or EOF.
+ // commonly used by state classes. It returns the corresponding "end-of"
+ // character in case it's a comment; otherwise it simply returns the
+ // current character.
+ int skipComment(int c) {
+ if (c == ';') {
+ while (true) {
+ c = source_->getChar();
+ if (c == '\n' || c == InputSource::END_OF_STREAM) {
+ return (c);
+ }
+ }
+ }
+ return (c);
+ }
std::vector<InputSourcePtr> sources_;
- Token token_;
+ InputSource* source_; // current source (NULL if sources_ is empty)
+ Token token_; // currently recognized token (set by a state)
+
+ // These are used in states, and defined here only as a placeholder.
+ // The main lexer class does not need these members.
+ size_t paren_count_; // nest count of the parentheses
+ bool last_was_eol_; // whether the lexer just passed an end-of-line
};
MasterLexer::MasterLexer() : impl_(new MasterLexerImpl) {
@@ -60,12 +85,14 @@ MasterLexer::pushSource(const char* filename, std::string* error) {
return (false);
}
+ impl_->source_ = impl_->sources_.back().get();
return (true);
}
void
MasterLexer::pushSource(std::istream& input) {
impl_->sources_.push_back(InputSourcePtr(new InputSource(input)));
+ impl_->source_ = impl_->sources_.back().get();
}
void
@@ -75,6 +102,8 @@ MasterLexer::popSource() {
"MasterLexer::popSource on an empty source");
}
impl_->sources_.pop_back();
+ impl_->source_ = impl_->sources_.empty() ? NULL :
+ impl_->sources_.back().get();
}
std::string
@@ -115,5 +144,144 @@ MasterLexer::Token::getErrorText() const {
return (error_text[val_.error_code_]);
}
+namespace master_lexer_internal {
+// Below we implement state classes for state transitions of MasterLexer.
+// Note that these need to be defined here so that they can refer to
+// the details of MasterLexerImpl.
+
+typedef MasterLexer::Token Token; // convenience shortcut
+
+bool
+State::wasLastEOL(const MasterLexer& lexer) const {
+ return (lexer.impl_->last_was_eol_);
+}
+
+const MasterLexer::Token&
+State::getToken(const MasterLexer& lexer) const {
+ return (lexer.impl_->token_);
+}
+
+size_t
+State::getParenCount(const MasterLexer& lexer) const {
+ return (lexer.impl_->paren_count_);
+}
+
+namespace {
+class CRLF : public State {
+public:
+ CRLF() {}
+ virtual ~CRLF() {} // see the base class for the destructor
+ virtual const State* handle(MasterLexer& lexer) const {
+ // We've just seen '\r'. If this is part of a sequence of '\r\n',
+ // we combine them as a single END-OF-LINE. Otherwise we treat the
+ // single '\r' as an EOL and continue tokeniziation from the character
+ // immediately after '\r'. One tricky case is that there's a comment
+ // between '\r' and '\n'. This implementation combines these
+ // characters and treats them as a single EOL (the behavior derived
+ // from BIND 9). Technically this may not be correct, but in practice
+ // the caller wouldn't distinguish this case from the case it has
+ // two EOLs, so we simplify the process.
+ const int c = getLexerImpl(lexer)->skipComment(
+ getLexerImpl(lexer)->source_->getChar());
+ if (c != '\n') {
+ getLexerImpl(lexer)->source_->ungetChar();
+ }
+ getLexerImpl(lexer)->token_ = Token(Token::END_OF_LINE);
+ getLexerImpl(lexer)->last_was_eol_ = true;
+ return (NULL);
+ }
+};
+
+// Currently this is provided mostly as a place holder
+class String : public State {
+public:
+ String() {}
+ virtual ~String() {} // see the base class for the destructor
+ virtual const State* handle(MasterLexer& /*lexer*/) const {
+ return (NULL);
+ }
+};
+
+// We use a common instance of a each state in a singleton-like way to save
+// construction overhead. They are not singletons in its strict sense as
+// we don't prohibit direct construction of these objects. But that doesn't
+// matter much anyway, because the definitions are completely hidden within
+// this file.
+const CRLF CRLF_STATE;
+const String STRING_STATE;
+}
+
+const State&
+State::getInstance(ID state_id) {
+ switch (state_id) {
+ case CRLF:
+ return (CRLF_STATE);
+ case String:
+ return (STRING_STATE);
+ }
+
+ // This is a bug of the caller, and this method is only expected to be
+ // used by tests, so we just forcefully make it fail by asserting the
+ // condition.
+ assert(false);
+ return (STRING_STATE); // a dummy return, to silence some compilers.
+}
+
+const State*
+State::start(MasterLexer& lexer, MasterLexer::Options options) {
+ // define some shortcuts
+ MasterLexer::MasterLexerImpl& lexerimpl = *lexer.impl_;
+ size_t& paren_count = lexerimpl.paren_count_;
+
+ while (true) {
+ const int c = lexerimpl.skipComment(lexerimpl.source_->getChar());
+ if (c == InputSource::END_OF_STREAM) {
+ lexerimpl.last_was_eol_ = false;
+ if (paren_count != 0) {
+ lexerimpl.token_ = Token(Token::UNBALANCED_PAREN);
+ paren_count = 0; // reset to 0; this helps in lenient mode.
+ return (NULL);
+ }
+ lexerimpl.token_ = Token(Token::END_OF_FILE);
+ return (NULL);
+ } else if (c == ' ' || c == '\t') {
+ // If requested and we are not in (), recognize the initial space.
+ if (lexerimpl.last_was_eol_ && paren_count == 0 &&
+ (options & MasterLexer::INITIAL_WS) != 0) {
+ lexerimpl.last_was_eol_ = false;
+ lexerimpl.token_ = Token(Token::INITIAL_WS);
+ return (NULL);
+ }
+ } else if (c == '\n') {
+ lexerimpl.last_was_eol_ = true;
+ if (paren_count == 0) { // we don't recognize EOL if we are in ()
+ lexerimpl.token_ = Token(Token::END_OF_LINE);
+ return (NULL);
+ }
+ } else if (c == '\r') {
+ if (paren_count == 0) { // check if we are in () (see above)
+ return (&CRLF_STATE);
+ }
+ } else if (c == '(') {
+ lexerimpl.last_was_eol_ = false;
+ ++paren_count;
+ } else if (c == ')') {
+ lexerimpl.last_was_eol_ = false;
+ if (paren_count == 0) {
+ lexerimpl.token_ = Token(Token::UNBALANCED_PAREN);
+ return (NULL);
+ }
+ --paren_count;
+ } else {
+ // Note: in #2373 we should probably ungetChar().
+ lexerimpl.last_was_eol_ = false;
+ return (&STRING_STATE);
+ }
+ // no code should be here; we just continue the loop.
+ }
+}
+
+} // namespace master_lexer_internal
+
} // end of namespace dns
} // end of namespace isc
diff --git a/src/lib/dns/master_lexer.h b/src/lib/dns/master_lexer.h
index da6bb5d..854d602 100644
--- a/src/lib/dns/master_lexer.h
+++ b/src/lib/dns/master_lexer.h
@@ -24,6 +24,9 @@
namespace isc {
namespace dns {
+namespace master_lexer_internal {
+class State;
+}
/// \brief Tokenizer for parsing DNS master files.
///
@@ -64,9 +67,22 @@ namespace dns {
/// this class does not throw for an error that would be reported as an
/// exception in other classes.
class MasterLexer {
+ friend class master_lexer_internal::State;
public:
class Token; // we define it separately for better readability
+ /// \brief Options for getNextToken.
+ ///
+ /// A compound option, indicating multiple options are set, can be
+ /// specified using the logical OR operator (operator|()).
+ enum Options {
+ NONE = 0, ///< No option
+ INITIAL_WS = 1, ///< recognize begin-of-line spaces after an
+ ///< end-of-line
+ QSTRING = 2, ///< recognize quoted string
+ NUMBER = 4 ///< recognize numeric text as integer
+ };
+
/// \brief The constructor.
///
/// \throw std::bad_alloc Internal resource allocation fails (rare case).
@@ -167,6 +183,16 @@ private:
MasterLexerImpl* impl_;
};
+/// \brief Operator to combine \c MasterLexer options
+///
+/// This is a trivial shortcut so that compound options can be specified
+/// in an intuitive way.
+inline MasterLexer::Options
+operator|(MasterLexer::Options o1, MasterLexer::Options o2) {
+ return (static_cast<MasterLexer::Options>(
+ static_cast<unsigned>(o1) | static_cast<unsigned>(o2)));
+}
+
/// \brief Tokens for \c MasterLexer
///
/// This is a simple value-class encapsulating a type of a lexer token and
@@ -192,7 +218,8 @@ public:
enum Type {
END_OF_LINE, ///< End of line detected (if asked for detecting it)
END_OF_FILE, ///< End of file detected (if asked for detecting it)
- INITIAL_WS, ///< White spaces at the beginning of a line
+ INITIAL_WS, ///< White spaces at the beginning of a line after an
+ ///< end of line
NOVALUE_TYPE_MAX = INITIAL_WS, ///< Max integer corresponding to
/// no-value (type only) types.
/// Mainly for internal use.
diff --git a/src/lib/dns/master_lexer_state.h b/src/lib/dns/master_lexer_state.h
new file mode 100644
index 0000000..1130f33
--- /dev/null
+++ b/src/lib/dns/master_lexer_state.h
@@ -0,0 +1,144 @@
+// Copyright (C) 2012 Internet Systems Consortium, Inc. ("ISC")
+//
+// Permission to use, copy, modify, and/or distribute this software for any
+// purpose with or without fee is hereby granted, provided that the above
+// copyright notice and this permission notice appear in all copies.
+//
+// THE SOFTWARE IS PROVIDED "AS IS" AND ISC DISCLAIMS ALL WARRANTIES WITH
+// REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY
+// AND FITNESS. IN NO EVENT SHALL ISC BE LIABLE FOR ANY SPECIAL, DIRECT,
+// INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM
+// LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE
+// OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
+// PERFORMANCE OF THIS SOFTWARE.
+
+#ifndef MASTER_LEXER_STATE_H
+#define MASTER_LEXER_STATE_H 1
+
+#include <dns/master_lexer.h>
+
+namespace isc {
+namespace dns {
+
+namespace master_lexer_internal {
+
+/// \brief Tokenization state for \c MasterLexer.
+///
+/// This is a base class of classes that represent various states of a single
+/// tokenization session of \c MasterLexer, i.e., the states used for a
+/// single call to \c MasterLexer::getNextToken().
+///
+/// It follows the convention of the state design pattern: each derived class
+/// corresponds to a specific state, and the state transition takes place
+/// through the virtual method named \c handle(). The \c handle() method
+/// takes the main \c MasterLexer object that holds all necessary internal
+/// context, and updates it as necessary; each \c State derived class is
+/// completely stateless.
+///
+/// The initial transition takes place in a static method of the base class,
+/// \c start(). This is mainly for implementation convenience; we need to
+/// pass options given to \c MasterLexer::getNextToken() for the initial
+/// state, so it makes more sense to separate the interface for the transition
+/// from the initial state.
+///
+/// When an object of a specific state class completes the session, it
+/// normally sets the identified token in the lexer, and returns NULL;
+/// if more transition is necessary, it returns a pointer to the next state
+/// object.
+///
+/// As is usual in the state design pattern, the \c State class is made
+/// a friend class of \c MasterLexer and can refer to its internal details.
+/// This is intentional; essentially its a part of \c MasterLexer and
+/// is defined as a separate class only for implementation clarity and better
+/// testability. It's defined in a publicly visible header, but that's only
+/// for testing purposes. No normal application or even no other classes of
+/// this library are expected to use this class.
+class State {
+public:
+ /// \brief Virtual destructor.
+ ///
+ /// In our usage this actually doesn't matter, but some compilers complain
+ /// about it and we need to silence them.
+ virtual ~State() {}
+
+ /// \brief Begin state transitions to get the next token.
+ ///
+ /// This is the first method that \c MasterLexer needs to call for a
+ /// tokenization session. The lexer passes a reference to itself
+ /// and options given in \c getNextToken().
+ ///
+ /// \throw InputSource::ReadError Unexpected I/O error
+ /// \throw std::bad_alloc Internal resource allocation failure
+ ///
+ /// \param lexer The lexer object that holds the main context.
+ /// \param options The options passed to getNextToken().
+ /// \return A pointer to the next state object or NULL if the transition
+ /// is completed.
+ static const State* start(MasterLexer& lexer,
+ MasterLexer::Options options);
+
+ /// \brief Handle the process of one specific state.
+ ///
+ /// This method is expected to be called on the object returned by
+ /// start(), and keep called on the returned object until NULL is
+ /// returned. The call chain will form the complete state transition.
+ ///
+ /// \throw InputSource::ReadError Unexpected I/O error
+ /// \throw std::bad_alloc Internal resource allocation failure
+ ///
+ /// \param lexer The lexer object that holds the main context.
+ /// \return A pointer to the next state object or NULL if the transition
+ /// is completed.
+ virtual const State* handle(MasterLexer& lexer) const = 0;
+
+ /// \brief Types of states.
+ ///
+ /// Specific states are basically hidden within the implementation,
+ /// but we'd like to allow tests to examine them, so we provide
+ /// a way to get an instance of a specific state.
+ enum ID {
+ CRLF, ///< Just seen a carriage-return character
+ String ///< Handling a string token
+ };
+
+ /// \brief Returns a \c State instance of the given state.
+ ///
+ /// This is provided only for testing purposes so tests can check
+ /// the behavior of each state separately. \c MasterLexer shouldn't
+ /// need this method.
+ static const State& getInstance(ID state_id);
+
+ /// \name Read-only accessors for testing purposes.
+ ///
+ /// These allow tests to inspect some selected portion of the internal
+ /// states of \c MasterLexer. These shouldn't be used except for testing
+ /// purposes.
+ ///@{
+ bool wasLastEOL(const MasterLexer& lexer) const;
+ const MasterLexer::Token& getToken(const MasterLexer& lexer) const;
+ size_t getParenCount(const MasterLexer& lexer) const;
+ ///@}
+
+protected:
+ /// \brief An accessor to the internal implementation class of
+ /// \c MasterLexer.
+ ///
+ /// This is provided for specific derived classes as they are not direct
+ /// friends of \c MasterLexer.
+ ///
+ /// \param lexer The lexer object that holds the main context.
+ /// \return A pointer to the implementation class object of the given
+ /// lexer. This is never NULL.
+ MasterLexer::MasterLexerImpl* getLexerImpl(MasterLexer& lexer) const {
+ return (lexer.impl_);
+ }
+};
+
+} // namespace master_lexer_internal
+} // namespace dns
+} // namespace isc
+#endif // MASTER_LEXER_STATE_H
+
+// Local Variables:
+// mode: c++
+// End:
diff --git a/src/lib/dns/tests/Makefile.am b/src/lib/dns/tests/Makefile.am
index d5adc21..33867da 100644
--- a/src/lib/dns/tests/Makefile.am
+++ b/src/lib/dns/tests/Makefile.am
@@ -27,6 +27,7 @@ run_unittests_SOURCES += labelsequence_unittest.cc
run_unittests_SOURCES += messagerenderer_unittest.cc
run_unittests_SOURCES += master_lexer_token_unittest.cc
run_unittests_SOURCES += master_lexer_unittest.cc
+run_unittests_SOURCES += master_lexer_state_unittest.cc
run_unittests_SOURCES += name_unittest.cc
run_unittests_SOURCES += nsec3hash_unittest.cc
run_unittests_SOURCES += rrclass_unittest.cc rrtype_unittest.cc
diff --git a/src/lib/dns/tests/master_lexer_state_unittest.cc b/src/lib/dns/tests/master_lexer_state_unittest.cc
new file mode 100644
index 0000000..bcee7fd
--- /dev/null
+++ b/src/lib/dns/tests/master_lexer_state_unittest.cc
@@ -0,0 +1,256 @@
+// Copyright (C) 2012 Internet Systems Consortium, Inc. ("ISC")
+//
+// Permission to use, copy, modify, and/or distribute this software for any
+// purpose with or without fee is hereby granted, provided that the above
+// copyright notice and this permission notice appear in all copies.
+//
+// THE SOFTWARE IS PROVIDED "AS IS" AND ISC DISCLAIMS ALL WARRANTIES WITH
+// REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY
+// AND FITNESS. IN NO EVENT SHALL ISC BE LIABLE FOR ANY SPECIAL, DIRECT,
+// INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM
+// LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE
+// OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
+// PERFORMANCE OF THIS SOFTWARE.
+
+#include <dns/master_lexer.h>
+#include <dns/master_lexer_inputsource.h>
+#include <dns/master_lexer_state.h>
+
+#include <gtest/gtest.h>
+
+#include <sstream>
+
+using namespace isc::dns;
+using namespace master_lexer_internal;
+
+namespace {
+typedef MasterLexer::Token Token; // shortcut
+
+class MasterLexerStateTest : public ::testing::Test {
+protected:
+ MasterLexerStateTest() : common_options(MasterLexer::INITIAL_WS),
+ s_null(NULL),
+ s_crlf(State::getInstance(State::CRLF)),
+ s_string(State::getInstance(State::String)),
+ options(MasterLexer::NONE),
+ orig_options(options)
+ {}
+
+ // Specify INITIAL_WS as common initial options.
+ const MasterLexer::Options common_options;
+ MasterLexer lexer;
+ const State* const s_null;
+ const State& s_crlf;
+ const State& s_string;
+ std::stringstream ss;
+ MasterLexer::Options options, orig_options;
+};
+
+// Common check for the end-of-file condition.
+// Token is set to END_OF_FILE, and the lexer was NOT last eol state.
+// Passed state can be any valid one; they are stateless, just providing the
+// interface for inspection.
+void
+eofCheck(const State& state, MasterLexer& lexer) {
+ EXPECT_EQ(Token::END_OF_FILE, state.getToken(lexer).getType());
+ EXPECT_FALSE(state.wasLastEOL(lexer));
+}
+
+TEST_F(MasterLexerStateTest, startAndEnd) {
+ // A simple case: the input is empty, so we begin with start and
+ // are immediately done.
+ lexer.pushSource(ss);
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ eofCheck(s_crlf, lexer);
+}
+
+TEST_F(MasterLexerStateTest, startToEOL) {
+ ss << "\n";
+ lexer.pushSource(ss);
+
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ EXPECT_TRUE(s_crlf.wasLastEOL(lexer));
+ EXPECT_EQ(Token::END_OF_LINE, s_crlf.getToken(lexer).getType());
+
+ // The next lexer session will reach EOF. Same eof check should pass.
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ eofCheck(s_crlf, lexer);
+}
+
+TEST_F(MasterLexerStateTest, space) {
+ // repeat '\t\n' twice (see below), then space after EOL
+ ss << " \t\n\t\n ";
+ lexer.pushSource(ss);
+
+ // by default space characters and tabs will be ignored. We check this
+ // twice; at the second iteration, it's a white space at the beginning
+ // of line, but since we don't specify INITIAL_WS option, it's treated as
+ // normal space and ignored.
+ for (size_t i = 0; i < 2; ++i) {
+ EXPECT_EQ(s_null, State::start(lexer, MasterLexer::NONE));
+ EXPECT_TRUE(s_crlf.wasLastEOL(lexer));
+ EXPECT_EQ(Token::END_OF_LINE, s_crlf.getToken(lexer).getType());
+ }
+
+ // Now we specify the INITIAL_WS option. It will be recognized and the
+ // corresponding token will be returned.
+ EXPECT_EQ(s_null, State::start(lexer, MasterLexer::INITIAL_WS));
+ EXPECT_FALSE(s_crlf.wasLastEOL(lexer));
+ EXPECT_EQ(Token::INITIAL_WS, s_crlf.getToken(lexer).getType());
+}
+
+TEST_F(MasterLexerStateTest, parentheses) {
+ ss << "\n(\na\n )\n "; // 1st \n is to check if 'was EOL' is set to false
+ lexer.pushSource(ss);
+
+ EXPECT_EQ(s_null, State::start(lexer, common_options)); // handle \n
+
+ // Now handle '('. It skips \n and recognize 'a' as string
+ EXPECT_EQ(0, s_crlf.getParenCount(lexer)); // check pre condition
+ EXPECT_EQ(&s_string, State::start(lexer, common_options));
+ EXPECT_EQ(1, s_crlf.getParenCount(lexer)); // check post condition
+ EXPECT_FALSE(s_crlf.wasLastEOL(lexer));
+
+ // skip 'a' (note: until #2373 it's actually skipped as part of the '('
+ // handling)
+ s_string.handle(lexer);
+
+ // Then handle ')'. '\n' before ')' isn't recognized because
+ // it's canceled due to the '('. Likewise, the space after the '\n'
+ // shouldn't be recognized but should be just ignored.
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ EXPECT_EQ(0, s_crlf.getParenCount(lexer));
+
+ // Now, temporarily disabled options are restored: Both EOL and the
+ // initial WS are recognized
+ EXPECT_EQ(Token::END_OF_LINE, s_crlf.getToken(lexer).getType());
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ EXPECT_EQ(Token::INITIAL_WS, s_crlf.getToken(lexer).getType());
+}
+
+TEST_F(MasterLexerStateTest, nestedParentheses) {
+ // This is an unusual, but allowed (in this implementation) case.
+ ss << "(a(b)\n c)\n ";
+ lexer.pushSource(ss);
+
+ EXPECT_EQ(&s_string, State::start(lexer, common_options)); // consume '('
+ s_string.handle(lexer); // consume 'a'
+ EXPECT_EQ(&s_string, State::start(lexer, common_options)); // consume '('
+ s_string.handle(lexer); // consume 'b'
+ EXPECT_EQ(2, s_crlf.getParenCount(lexer)); // now the count is 2
+
+ // Close the inner most parentheses. count will be decreased, but option
+ // shouldn't be restored yet, so the intermediate EOL or initial WS won't
+ // be recognized.
+ EXPECT_EQ(&s_string, State::start(lexer, common_options)); // consume ')'
+ s_string.handle(lexer); // consume 'c'
+ EXPECT_EQ(1, s_crlf.getParenCount(lexer));
+
+ // Close the outermost parentheses. count will be reset to 0, and original
+ // options are restored.
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+
+ // Now, temporarily disabled options are restored: Both EOL and the
+ // initial WS are recognized
+ EXPECT_EQ(Token::END_OF_LINE, s_crlf.getToken(lexer).getType());
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ EXPECT_EQ(Token::INITIAL_WS, s_crlf.getToken(lexer).getType());
+}
+
+TEST_F(MasterLexerStateTest, unbalancedParentheses) {
+ // Only closing paren is provided. We prepend a \n to check if it's
+ // correctly canceled after detecting the error.
+ ss << "\n)";
+ ss << "(a";
+ lexer.pushSource(ss);
+
+ EXPECT_EQ(s_null, State::start(lexer, common_options)); // consume '\n'
+ EXPECT_TRUE(s_crlf.wasLastEOL(lexer)); // this \n was remembered
+
+ // Now checking ')'. The result should be error, count shouldn't be
+ // changed. "last EOL" should be canceled.
+ EXPECT_EQ(0, s_crlf.getParenCount(lexer));
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ EXPECT_EQ(0, s_crlf.getParenCount(lexer));
+ ASSERT_EQ(Token::ERROR, s_crlf.getToken(lexer).getType());
+ EXPECT_EQ(Token::UNBALANCED_PAREN, s_crlf.getToken(lexer).getErrorCode());
+ EXPECT_FALSE(s_crlf.wasLastEOL(lexer));
+
+ // Reach EOF with a dangling open parenthesis.
+ EXPECT_EQ(&s_string, State::start(lexer, common_options)); // consume '('
+ s_string.handle(lexer); // consume 'a'
+ EXPECT_EQ(1, s_crlf.getParenCount(lexer));
+ EXPECT_EQ(s_null, State::start(lexer, common_options)); // reach EOF
+ ASSERT_EQ(Token::ERROR, s_crlf.getToken(lexer).getType());
+ EXPECT_EQ(Token::UNBALANCED_PAREN, s_crlf.getToken(lexer).getErrorCode());
+ EXPECT_EQ(0, s_crlf.getParenCount(lexer)); // should be reset to 0
+}
+
+TEST_F(MasterLexerStateTest, startToComment) {
+ // Begin with 'start', skip space, then encounter a comment. Skip
+ // the rest of the line, and recognize the new line. Note that the
+ // second ';' is simply ignored.
+ ss << " ;a;\n";
+ ss << ";a;"; // Likewise, but the comment ends with EOF.
+ lexer.pushSource(ss);
+
+ // Comment ending with EOL
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ EXPECT_EQ(Token::END_OF_LINE, s_crlf.getToken(lexer).getType());
+
+ // Comment ending with EOF
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ EXPECT_EQ(Token::END_OF_FILE, s_crlf.getToken(lexer).getType());
+}
+
+TEST_F(MasterLexerStateTest, commentAfterParen) {
+ // comment after an opening parenthesis. The code that is tested by
+ // other tests should also ensure that it works correctly, but we
+ // check it explicitly.
+ ss << "( ;this is a comment\na)\n";
+ lexer.pushSource(ss);
+
+ // consume '(', skip comments, consume 'a', then consume ')'
+ EXPECT_EQ(&s_string, State::start(lexer, common_options));
+ s_string.handle(lexer);
+ EXPECT_EQ(s_null, State::start(lexer, common_options));
+ EXPECT_EQ(Token::END_OF_LINE, s_crlf.getToken(lexer).getType());
+}
+
+TEST_F(MasterLexerStateTest, crlf) {
+ ss << "\r\n"; // case 1
+ ss << "\r "; // case 2
+ ss << "\r;comment\na"; // case 3
+ ss << "\r"; // case 4
+ lexer.pushSource(ss);
+
+ // 1. A sequence of \r, \n is recognized as a single 'end-of-line'
+ EXPECT_EQ(&s_crlf, State::start(lexer, common_options)); // recognize '\r'
+ EXPECT_EQ(s_null, s_crlf.handle(lexer)); // recognize '\n'
+ EXPECT_EQ(Token::END_OF_LINE, s_crlf.getToken(lexer).getType());
+ EXPECT_TRUE(s_crlf.wasLastEOL(lexer));
+
+ // 2. Single '\r' (not followed by \n) is recognized as a single
+ // 'end-of-line'. then there will be "initial WS"
+ EXPECT_EQ(&s_crlf, State::start(lexer, common_options)); // recognize '\r'
+ // see ' ', "unget" it
+ EXPECT_EQ(s_null, s_crlf.handle(lexer));
+ EXPECT_EQ(s_null, State::start(lexer, common_options)); // recognize ' '
+ EXPECT_EQ(Token::INITIAL_WS, s_crlf.getToken(lexer).getType());
+
+ // 3. comment between \r and \n
+ EXPECT_EQ(&s_crlf, State::start(lexer, common_options)); // recognize '\r'
+ // skip comments, recognize '\n'
+ EXPECT_EQ(s_null, s_crlf.handle(lexer));
+ EXPECT_EQ(Token::END_OF_LINE, s_crlf.getToken(lexer).getType());
+ EXPECT_EQ(&s_string, State::start(lexer, common_options));
+
+ // 4. \r then EOF
+ EXPECT_EQ(&s_crlf, State::start(lexer, common_options)); // recognize '\r'
+ // see EOF, then "unget" it
+ EXPECT_EQ(s_null, s_crlf.handle(lexer));
+ EXPECT_EQ(s_null, State::start(lexer, common_options)); // recognize EOF
+ EXPECT_EQ(Token::END_OF_FILE, s_crlf.getToken(lexer).getType());
+}
+
+}
diff --git a/tests/tools/perfdhcp/templates/Makefile.am b/tests/tools/perfdhcp/templates/Makefile.am
index 33930a7..c22787f 100644
--- a/tests/tools/perfdhcp/templates/Makefile.am
+++ b/tests/tools/perfdhcp/templates/Makefile.am
@@ -5,8 +5,6 @@ SUBDIRS = .
CLEANFILES = test1.hex test2.hex test3.hex test4.hex test5.hex
perfdhcpdir = $(pkgdatadir)
-perfdhcp_DATA = discover-example.hex request4-example.hex
-perfdhcp_DATA += solicit-example.hex request6-example.hex
EXTRA_DIST = discover-example.hex request4-example.hex
EXTRA_DIST += solicit-example.hex request6-example.hex
More information about the bind10-changes
mailing list