Simon Michael
b6ff170688
lib: simplify format detection, avoid ledger reader by default
...
When we don't know a file's format, instead of choosing a subset of
readers based on content sniffing, now we just try them all.
Also, LedgerReader is now used only as a last resort,
as it's not yet competitive with JournalReader.
2016-11-18 13:24:57 -08:00
Simon Michael
3ddc9d7432
lib: clarify file format detectors
2016-11-16 13:25:33 -08:00
Simon Michael
4ceda8f985
timeclock: fix order of transaction indexes #394
...
Timeclock transaction ids now count up rather than down.
Also, remove old code for appending timeclock transactions to journal transactions,
a holdover from the days when both were allowed in one file.
2016-08-14 14:50:03 -07:00
Moritz Kiefer
4141067428
Replace Parsec with Megaparsec (see #289 ) ( #366 )
...
* Replace Parsec with Megaparsec (see #289 )
This builds upon PR #289 by @rasendubi
* Revert renaming of parseWithState to parseWithCtx
* Fix doctests
* Update for Megaparsec 5
* Specialize parser to improve performance
* Pretty print errors
* Swap StateT and ParsecT
This is necessary to get the correct backtracking behavior, i.e. discard
state changes if the parsing fails.
2016-07-29 08:57:10 -07:00
Simon Michael
90c9735b7a
lib: textification: descriptions & codes
...
Slightly higher (with small files) and lower (with large files) maximum
residency, and slightly quicker for all.
hledger -f data/100x100x10.journal stats
<<ghc: 42858472 bytes, 84 GCs, 193712/269608 avg/max bytes residency (3 samples), 2M in use, 0.000 INIT (0.015 elapsed), 0.016 MUT (0.042 elapsed), 0.011 GC (0.119 elapsed) :ghc>>
<<ghc: 42891776 bytes, 84 GCs, 190816/260920 avg/max bytes residency (3 samples), 2M in use, 0.000 INIT (0.004 elapsed), 0.017 MUT (0.025 elapsed), 0.010 GC (0.015 elapsed) :ghc>>
hledger -f data/1000x1000x10.journal stats
<<ghc: 349575240 bytes, 681 GCs, 1396425/4091680 avg/max bytes residency (7 samples), 11M in use, 0.000 INIT (0.000 elapsed), 0.137 MUT (0.146 elapsed), 0.050 GC (0.057 elapsed) :ghc>>
<<ghc: 349927568 bytes, 681 GCs, 1397825/4097248 avg/max bytes residency (7 samples), 11M in use, 0.000 INIT (0.000 elapsed), 0.126 MUT (0.133 elapsed), 0.050 GC (0.057 elapsed) :ghc>>
hledger -f data/10000x1000x10.journal stats
<<ghc: 3424029496 bytes, 6658 GCs, 11403141/41077288 avg/max bytes residency (11 samples), 111M in use, 0.000 INIT (0.000 elapsed), 1.278 MUT (1.310 elapsed), 0.493 GC (0.546 elapsed) :ghc>>
<<ghc: 3427418064 bytes, 6665 GCs, 11127869/37790168 avg/max bytes residency (11 samples), 109M in use, 0.000 INIT (0.001 elapsed), 1.212 MUT (1.229 elapsed), 0.466 GC (0.519 elapsed) :ghc>>
hledger -f data/100000x1000x10.journal stats
<<ghc: 34306546248 bytes, 66727 GCs, 77030638/414617944 avg/max bytes residency (14 samples), 1012M in use, 0.000 INIT (0.000 elapsed), 12.965 MUT (13.164 elapsed), 4.771 GC (5.447 elapsed) :ghc>>
<<ghc: 34340246056 bytes, 66779 GCs, 76983178/416011480 avg/max bytes residency (14 samples), 1011M in use, 0.000 INIT (0.008 elapsed), 12.666 MUT (12.836 elapsed), 4.595 GC (5.175 elapsed) :ghc>>
2016-05-24 19:00:58 -07:00
Simon Michael
c89c33b36e
lib: textification: parse stream
...
10% more allocation, but 35% lower maximum residency, and slightly quicker.
hledger -f data/100x100x10.journal stats
<<ghc: 39327768 bytes, 77 GCs, 196834/269496 avg/max bytes residency (3 samples), 2M in use, 0.000 INIT (0.010 elapsed), 0.020 MUT (0.092 elapsed), 0.014 GC (0.119 elapsed) :ghc>>
<<ghc: 42842136 bytes, 84 GCs, 194010/270912 avg/max bytes residency (3 samples), 2M in use, 0.000 INIT (0.009 elapsed), 0.016 MUT (0.029 elapsed), 0.012 GC (0.120 elapsed) :ghc>>
hledger -f data/1000x1000x10.journal stats
<<ghc: 314291440 bytes, 612 GCs, 2070776/6628048 avg/max bytes residency (7 samples), 16M in use, 0.000 INIT (0.000 elapsed), 0.128 MUT (0.144 elapsed), 0.059 GC (0.070 elapsed) :ghc>>
<<ghc: 349558872 bytes, 681 GCs, 1397597/4106384 avg/max bytes residency (7 samples), 11M in use, 0.000 INIT (0.004 elapsed), 0.124 MUT (0.133 elapsed), 0.047 GC (0.053 elapsed) :ghc>>
hledger -f data/10000x1000x10.journal stats
<<ghc: 3070026824 bytes, 5973 GCs, 12698030/62951784 avg/max bytes residency (10 samples), 124M in use, 0.000 INIT (0.002 elapsed), 1.268 MUT (1.354 elapsed), 0.514 GC (0.587 elapsed) :ghc>>
<<ghc: 3424013128 bytes, 6658 GCs, 11405501/41071624 avg/max bytes residency (11 samples), 111M in use, 0.000 INIT (0.001 elapsed), 1.343 MUT (1.406 elapsed), 0.511 GC (0.573 elapsed) :ghc>>
hledger -f data/100000x1000x10.journal stats
<<ghc: 30753387392 bytes, 59811 GCs, 117615462/666703600 avg/max bytes residency (14 samples), 1588M in use, 0.000 INIT (0.000 elapsed), 12.068 MUT (12.238 elapsed), 6.015 GC (7.190 elapsed) :ghc>>
<<ghc: 34306530696 bytes, 66727 GCs, 76806196/414629312 avg/max bytes residency (14 samples), 1009M in use, 0.000 INIT (0.010 elapsed), 14.357 MUT (16.370 elapsed), 5.298 GC (6.534 elapsed) :ghc>>
2016-05-24 19:00:57 -07:00
Simon Michael
2538d14ea7
lib: textification begins! account names
...
The first of several conversions from String to (strict) Text, hopefully
reducing space and time usage.
This one shows a small improvement, with GHC 7.10.3 and text-1.2.2.1:
hledger -f data/100x100x10.journal stats
string: <<ghc: 39471064 bytes, 77 GCs, 198421/275048 avg/max bytes residency (3 samples), 2M in use, 0.000 INIT (0.001 elapsed), 0.015 MUT (0.020 elapsed), 0.010 GC (0.014 elapsed) :ghc>>
text: <<ghc: 39268024 bytes, 77 GCs, 197018/270840 avg/max bytes residency (3 samples), 2M in use, 0.000 INIT (0.002 elapsed), 0.016 MUT (0.022 elapsed), 0.009 GC (0.011 elapsed) :ghc>>
hledger -f data/1000x100x10.journal stats
string: <<ghc: 318555920 bytes, 617 GCs, 2178997/7134472 avg/max bytes residency (7 samples), 16M in use, 0.000 INIT (0.001 elapsed), 0.129 MUT (0.136 elapsed), 0.067 GC (0.077 elapsed) :ghc>>
text: <<ghc: 314248496 bytes, 612 GCs, 2074045/6617960 avg/max bytes residency (7 samples), 16M in use, 0.000 INIT (0.003 elapsed), 0.137 MUT (0.145 elapsed), 0.067 GC (0.079 elapsed) :ghc>>
hledger -f data/10000x100x10.journal stats
string: <<ghc: 3114763608 bytes, 6026 GCs, 18858950/75552024 avg/max bytes residency (11 samples), 201M in use, 0.000 INIT (0.000 elapsed), 1.331 MUT (1.372 elapsed), 0.699 GC (0.812 elapsed) :ghc>>
text: <<ghc: 3071468920 bytes, 5968 GCs, 14120344/62951360 avg/max bytes residency (9 samples), 124M in use, 0.000 INIT (0.003 elapsed), 1.272 MUT (1.349 elapsed), 0.513 GC (0.578 elapsed) :ghc>>
hledger -f data/100000x100x10.journal stats
string: <<ghc: 31186579432 bytes, 60278 GCs, 135332581/740228992 avg/max bytes residency (13 samples), 1697M in use, 0.000 INIT (0.008 elapsed), 14.677 MUT (15.508 elapsed), 7.081 GC (8.074 elapsed) :ghc>>
text: <<ghc: 30753427672 bytes, 59763 GCs, 117595958/666457240 avg/max bytes residency (14 samples), 1588M in use, 0.000 INIT (0.008 elapsed), 13.713 MUT (13.966 elapsed), 6.220 GC (7.108 elapsed) :ghc>>
2016-05-24 19:00:49 -07:00
Simon Michael
0f5ee154c4
lib: simplify parsers; cleanups ( #275 )
...
The journal/timeclock/timedot parsers, instead of constructing (opaque)
journal update functions which are later applied to build the journal,
now construct the journal directly (by modifying the parser state). This
is easier to understand and debug. It also removes any possibility of
the journal updates being a space leak. (They weren't, in fact memory
usage is now slightly higher, but that will be addressed in other ways.)
Also:
Journal data and journal parse info have been merged into one type (for
now), and field names are more consistent.
The ParsedJournal type alias has been added to distinguish being-parsed
and finalised journals.
Journal is now a monoid.
stats: fixed an issue with ordering of include files
journal: fixed an issue with ordering of included same-date transactions
timeclock: sessions can no longer span file boundaries (unclocked-out
sessions will be auto-closed at the end of the file).
expandPath now throws a proper IO error (and requires the IO monad).
2016-05-23 00:44:19 -07:00
Simon Michael
7f5e09096f
lib: rename JournalContext to JournalParseState
2016-05-18 20:57:34 -07:00
Simon Michael
84097b75c7
journal: can now include timeclock/timedot files ( #320 )
...
journal files can now include journal, timeclock or timedot files (but
not yet CSV files). Also timeclock/timedot files no longer support
default year directives.
The Hledger.Read.* modules have been reorganised for better reuse.
Hledger.Read.Utils has been renamed Hledger.Read.Common and holds
low-level parsers & utilities; high-level read utilities have moved to
Hledger.Read.
2016-05-17 19:46:54 -07:00
Simon Michael
588f36d662
rename timelog format to timeclock
2016-04-12 21:13:19 -07:00