README.md
# tempsdb
[](https://pypi.python.org/pypi/tempsdb)
[](https://badge.fury.io/py/tempsdb)
[](https://pypi.python.org/pypi/tempsdb)
[](http://tempsdb.readthedocs.io/en/latest/?badge=latest)
[](https://codeclimate.com/github/smok-serwis/tempsdb/maintainability)
[](https://codeclimate.com/github/smok-serwis/tempsdb/test_coverage)
[](https://travis-ci.com/smok-serwis/tempsdb)
[](https://pypi.org/project/tempsdb/)
[](https://github.com/smok-serwis/tempsdb)
Embedded Cython library for time series that you need to upload somewhere.
Stored time series with a 8-byte timestamp and a data, which can be of
fixed length or variable.
# Installation
```bash
git clone https://github.com/smok-serwis/tempsdb
cd tempsdb
pip install snakehouse tempsdb
python setup.py install
```
You need both [snakehouse](https://pypi.org/project/snakehouse/1.2.2/)
and [tempsdb](https://pypi.org/project/tempsdb/) to compile it from the source,
though binary wheels are available for whatever you'd like.
If there's a binary wheel that you would like to have,
just drop me an [issue](https://github.com/smok-serwis/tempsdb/issues/new)/
If you're installing it somewhere that you don't need both snakehouse
and tempsdb installed, compile your own binary wheel with
```
python setup.py bdist_wheel
```
Then copy your resulting wheel and install it via pip on the target system.
Be aware that tempsdb does logging.
Consult the docs for how to disable it.
# Changelog
## v0.6.5
* works on Python 3.10 and 3.11
* fixed warnings about signed to unsigned comparison
## v0.6.4
* fixed a bug with slicing chunks in `VarlenSeries`
* added extra comparison operators for `VarlenEntry`
* added `sync` to `VarlenSeries`
* fixed a bug with not propagating metadata write exceptions
* fixed a bug with `Database` treating `varlen` and metadata as real time series
## v0.6.3
* added logging for opening and closing series
## v0.6.2
* added the context manager syntax to VarlenIterator
* fixed a memory leak that happened during getting current value
from an empty series
* added `MemoryPressureManager` support for `Database`
## v0.6.1
* fixed an issue with `Iterators`
## v0.6
* **bugfix**: fixed some bugs with reading values after close
* added support for storing metadata as minijson
* this will be enabled by default is minijson is importable
* fixed minor compiler warnings
* `TimeSeries.iterate_range` will accept a parameter called
`direct_bytes` for compatibility with `VarlenSeries`.
It's value is ignored
* more class constructors use explicit typing - faster tempsdb
* `TimeSeries.get_current_value` will correctly raise `ValueError` instead of returning None
## v0.5.4
* older TempsDB databases that do not support varlens will be updated upon opening
* added metadata support for databases
* a flush will be done before re-enabling mmap
* bugfix to read archive data
## v0.5.3
* added `disable_mmap`, `enable_mmap` and `open_chunks_mmap_size` into `VarlenSeries`
## v0.5.2
* added multiple properties and attributes to `VarlenSeries`
## v0.5.1
* added `VarlenSeries.close_chunks`
* `Database.sync` will now return 0
* indexed-gzip proved to be a poor choice, dropped
* `setup.py` fixed
## v0.5
* if mmap is used, the kernel will be informed after loading the chunk that we
don't need it's memory right now
* deleting a `TimeSeries` will now correctly return a zero
* both `Database`, `TimeSeries` and `Chunk` destructor will close and
emit a warning if the user forgot to
* if page_size is default, it won't be written as part of the metadata
* added support for per-series metadata
* following additions to `Database`:
* `delete_series`
* `delete_varlen_series`
* following additions to `TimeSeries`:
* added `append_padded`
* added metadata support, `metadata` property and `set_metadata` call
* added variable length series
* added experimental support for gzipping time series
* fixed a bug where getting a series that was already closed would TypeError
* following additions to `Chunk`:
* `get_slice_of_piece_at`
* `get_slice_of_piece_starting_at`
* `get_byte_of_piece`
* `get_timestamp_at`
* fixed the behaviour of `AlternativeMMaps` when passed a single index to __getitem__ and __setitem__
* added `StillOpen` exception, chunk won't allow to close itself if it has any
remaining references
## v0.4.4
* more error conditions during mmap will be supported as well
* ENOMEM will be correctly handled during resize operation
* added `TimeSeries.descriptor_based_access`
* added `Chunk.switch_to_mmap_based_access`
## v0.4.3
* improving handling mmap failures on too low memory
* slightly reduced `metadata.txt` by defaulting `page_size`
* moved `Chunk`
* added support for gzipping
* added `DirectChunk`
* iterating and writing at the same time from multiple threads
made safe
* added `TimeSeries.disable_mmap`
* `Iterator`'s destructor will emit a warning if you forget to close it explicitly.
* added option for transparent gzip compression
Please note that gzip disables mmap!
* experimental gzip support for constant-length time series
## v0.4.2
* empty series will return an Iterator
* **bugfix release** fixed `Database.create_series`
* `Database` constructor will throw if no database is there
* changed `Iterator.next` to `Iterator.next_item`,
synce Cython guys said to not implement the method `next`
on iterators.
## v0.4.1
* **bugfix release** fixed `get_open_series`
## v0.4
* can install from sdist now
## v0.3
* added `TimeSeries.get_current_value`
* added `Database.sync`
## v0.2
* added `get_open_series`
* added `get_all_series`
* added `get_first_entry_for`
* added `close_all_open_series`
* added `TimeSeries.name`
* added option to use descriptor based access instead of mmap
* added `TimeSeries.open_chunks_ram_size`
## v0.1
First release