LibUV event will use the best available polling method on a system, kind
of like aria2 does already with the different *EventPoll
implementations.
However, libuv may support different/newer polling mechanisms; for
example on Windows it will use IO Completion Ports which are superior to
select() ;)
Now dedicated ConnectCommand handles connection establishment. It
checks whether connection is established or not. It also handles
backup connection. The next Command creation is abstracted using
ControlChain struct template.
Saved sessions may very large, as in hundreds and even thousands of
megabyte when dealing with large queues.
Add support to save and reload sessions to gzipped files, when libz is
available.
The session serializer will output gzipped contents when the file ends
with .gz, while the input file reader (UriListParser) will accept
whatever is thrown at it.
--save-session-interval option saves error/unfinished downloads to a
file specified by --save-session option every SEC seconds. If 0 is
given, file will be saved only when aria2 exits.
This change replaces the current 64 bit sequential GID with 64 bits
random bytes GID in an attempt to support persistent GID. Internally,
the GID is stored as uint64_t. For human representation and RPC
interface, GID is represented as 16 bytes hex string. For console
readout, 16 bytes are too long, so it is abbreviated to first 6 bytes.
When querying GID in RPC calls, user can speicfy the prefix of GID as
long as the prefix is shared by more than 1 GID entries.
This option enables disk cache. If SIZE is 0, the disk cache is
disabled. This feature caches the downloaded data in memory, which
grows to at most SIZE bytes. The cache storage is created for aria2
instance and shared by all downloads. The one advantage of the disk
cache is reduce the disk seek time because the data is written in
larger unit and it is reordered by the offset of the file. If the
underlying file is heavily fragmented it is not the case.
The old implementation calculates download/upload statistics for a
RequestGroup by summing up all PeerStat objects. For global
statistics, those are summed together. This clearly incurs runtime
penalty and we introduced some kind of caching which updates
statistics every 250ms but it did not work right.
This change removes all these aggregation code, and instead makes
RequestGroup and RequestGroupMan objects hold NetStat object and
download/upload bytes are directly calculated by thier own NetStat.
This is far more simplar than the old way and less runtime penalty and
brings more accuracy.
--file-allocation option can now take new value 'trunc'. 'trunc' uses
ftruncate() system call or platform-specific counterpart to truncate a
file to a specified length.
The help messages shown when argument error were redesigned and less
verbose now. When unknown or ambiguous option is given, show
suggestions like "Did you mean...". Some constant values related to
levenstein distance are borrowed from git help.c.
This feature stops application when process PID is not running. This
is useful if aria2 process is forked from a parent process. The parent
process can fork aria2 with its own pid and when parent process exits
for some reason, aria2 can detect it and shutdown itself.
The code which detects whether or not given process PID is running is
contributed by Emmanuel Engelhart.
libnettle and libgmp is used by default. Old implementation uses
libgcrypt as default. If more and more Linux distributions offer
libnettle linked GnuTLS, then this change decrease library dependency
against libgcrypt and libgpg-error. The current library detection for
libnettle and libgmp is very simple, just use AC_SEARCH_LIBS.
util::generateRandomData() was written using SimpleRandomizer, thus no
external library dependency from this function.
In MinGW32 build, global::cout is WinConsoleFile. When outputing to
console, it uses WriteConsoleW, so Unicode characters are written.
Fixed memory leak in WChar<->MultiByte conversion functions. The
portion of code which uses std::cout is rewritten to use global::cout.
This option specifies piece selection algorithm used in HTTP/FTP
download. Piece means fixed length segment which is downloaded in
parallel in segmented download. If 'default' is given, aria2 selects
piece so that it reduces the number of establishing connection. This
is reasonable default behaviour because establishing connection is an
expensive operation. If 'inorder' is given, aria2 selects piece which
has minimum index. Index=0 means first of the file. This will be
useful to view movie while downloading it. --enable-http-pipelining
option may be useful to reduce reconnection overhead. Please note
that aria2 honors --min-split-size option, so it will be necessary to
specify a reasonable value to --min-split-size option.
We introduced SocketRecvBuffer which buffers received bytes. Since
HTTP response header and response body are divided with \r\n, we have
to buffer up several bytes to find this delimiter. We use
SocketRecvBuffer to hold these bytes and only consumes header and
passes SocketRecvBuffer, which may contain head of response body, to
next Command. Since FTPConnection doesn't use SocketCore::peekData(),
we left it as is.
We added 2 macros A2_TEST_DIR and A2_TEST_OUT_DIR to pass tests in
`make distcheck`. A2_TEST_DIR refers to test directory. All output
files by unit tests are now created under A2_TEST_OUT_DIR directory.
Added .cc file for classes/structs that only provided by header
file. Defined non-POD classes' ctor, dtor in .cc file. Moved
implementation code in header file to .cc file for major
classes/strucsts.
Sort usedHosts by least used and faster download speed.
* src/Makefile.am
* src/RequestGroupMan.cc
* src/Triplet.h
* test/Makefile.am
* test/TripletTest.cc
Added keys parameter to aria2.tellStatus, aria2.tellActive,
aria2.tellWaiting and aria2.tellStopped XML-RPC method. 'keys' is
array of string. If it is specified, the response contains only
keys in 'keys' array. If 'keys' is empty or not specified, the
response contains all keys. This is useful when you just want
specific keys and avoid unnecessary transfers. For example,
*aria2.tellStatus*("1", ["gid", "status"]) returns 'gid' and
'status' key. Made get*Param() functions XmlRpcRequest's
methods and changed portions of the code that were affected by
this change.
* doc/aria2c.1.txt
* src/Makefile.am
* src/XmlRpcMethodImpl.cc
* src/XmlRpcMethodImpl.h
* src/XmlRpcRequest.cc
* src/XmlRpcRequest.h
* test/XmlRpcMethodTest.cc
Moved URI parser code to uri.h, uri.cc to provide fast URI parser
without Request object initialization.
* src/Makefile.am
* src/Request.cc
* src/uri.cc
* src/uri.h
* test/Makefile.am
* test/RequestTest.cc
* test/UriTest.cc
When allocating disk space, for Linux system with fallocate()
system call, first check file system supports fallocate. This
just run fallocate with small chunk and see it succeeds or fails.
If it succeeds, use fallocate() to allocate entire file otherwise
fall back to traditional slower method: writing zeros. This
behavior is enabled in --file-allocation=prealloc, so this is
enabled by default for most modern Linux.
* configure.ac
* src/AbstractDiskWriter.cc
* src/AbstractDiskWriter.h
* src/AbstractSingleDiskAdaptor.cc
* src/AdaptiveFileAllocationIterator.cc
* src/AdaptiveFileAllocationIterator.h
* src/DefaultPieceStorage.cc
* src/DiskAdaptor.cc
* src/DiskAdaptor.h
* src/FallocFileAllocationIterator.cc
* src/Makefile.am
* src/MultiFileAllocationIterator.cc
* src/OptionHandlerFactory.cc
* test/FallocFileAllocationIteratorTest.cc
* test/Makefile.am
Supported poll() for socket event notification. --event-poll can
take value "poll". Fixed the bug that timeout for
SocketCore::isReadable()/isWritable() is ignored when epoll is
used.
* src/OptionHandlerFactory.cc
* src/a2io.h
* src/prefs.h
* src/SocketCore.h
* src/Makefile.am
* src/main.cc
* src/PollEventPoll.h
* src/SocketCore.cc
* src/DownloadEngineFactory.cc
* src/PollEventPoll.cc
* src/prefs.cc
* configure.ac
Implemented clock_gettime() using mach_absolute_time in Mac OS X.
* configure.ac
* src/Makefile.am
* src/a2time.h
* src/clock_gettime_osx.cc
* src/clock_gettime_osx.h
Nils Maier