## v0.5.1
### Features
* Generate and use `csvpo` cache files (increase compilation speed)
* Ugprade `umya-spreadsheet` to v1.1.1
* Deprecate `Error::ObjectCodeError`
### Bugfixes
* Fix leaking of namespaces from dependencies making their way into the dependent. Only exported
symbols should have been propagated
* Fix evaluation of variables dependent on variables from another file
* Fix lexing of the `fn` keyword in places like `fn fn_name` where the "fn" part is repeated twice
## v0.5.0
### Features
* `use` statement for importing code from other files
* Using object code files (.csvpo) during compilation
* Allow for `.` characters in a function name
* Support various levels of verbosity by repeating -v (i.e., -vvvv)
* Improved error messages for syntax errors in cells
* Use a proper logger (env\_logger) and tune output according to the -v[vvv] flag
* Code coverage reporting
* Tooling to run benchmarks
### Bugfixes
* Fix unfriendly error when calling a function with the wrong number of arguments
### **Breaking (Language) Changes**
* All builtin variables & functions are removed. All of the functionality provided by them could
be done with native spreadsheet functions anyway. And we're adding support for module loading
which means they could just be implemented as a lib instead.
* Rename `![[expand]]` to `![[fill]]` and all related code references
* Rename `[[format]]` to `[[text]]`
* Rename `Template` to `Module`
* Rename `CodeSection` to `Scope`
* Rename `Error::ObjectWriteError` to `Error::ObjectCodeError`
* `Template.write_object_file` made crate-private
* `Error::EvalError` changed
## v0.4.1
### Features
* data validations via the `validate=...` syntax
* Google Sheets API access via user accounts (which previously required service accounts)
* More clear and colored error messages throughout
* Dates assume local TZ if not specified
### Bugfixes
* Variables defined on rows (i.e., `![[var=this_row]],=this_row`) weren't being evaluated to the
correct results
* Expands were incorrectly filling into the rows below them, only for excel
* More consistent parsing of dates
* Fix line count in syntax error messages
### **Breaking Changes**
* Various parser classes made crate-private
## v0.4.0
### Features
* Ability to reference variables defined in an expand, outside of an expand. If they're
referenced inside an expand they'll resolve to their exact location, otherwise they resolve to
the range represented by that column in the range.
* Support for `![[var=...]]` both in and outside expands. They'll reference either a row (if
defined outside an expand), a row relative to the expand (if defined and referenced in an expand)
or the entire row range of the expand (if defined in an expand and referenced outside it).
* Much better error reporting - pretty much everything now includes some contextual code highlighting
* Excel: `note` support
* More useful `-v/--verbose` output
### Bugfixes
* Fix cellabove/cellbelow/celladjacent to all take columns (A/B/C/etc)
* Reading and writing CSV with inconsistent column lengths would fail
* Trim leading and trailing spaces on input CSV
* Excel: Fix workbooks being created with an empty first sheet
* Excel: Fix background/foreground coloring
### **Breaking Changes**
* A lot of functions and structs made `pub(crate)` privacy
* Error classes significantly refactored
## v0.3.0
Complete rewrite in Rust, which also includes:
* shorthand notation for modifiers
* much speed
* a re-usable A1 library (a1\_notation)
* better error messages since I've hand-written the parser
## v0.2.1
- Add a `-s`/`--safe` flag which changes the merge strategy to not overwrite existing values. If
the spreadsheet being written to has values where the csvpp template wants to write, they will
be overwritten otherwise.
## v0.2.0
### **Breaking Changes**
- Removal of the $$ operator - to dereference variables you can just reference them by name and
they will be resolved if they are defined. Otherwise they will be left alone in the output
### Non-breaking Changes
- Excel: fix the merging of existing values
- CSV: fix the merging of existing values
- Support merging in values from CSV (previously it would ignore/overwrite them)
- Allow for more generous spacing in the csv section (and reflect this in the examples)
- More type coverage
## v0.1.3
- Proper scoping of variables defined within an expand modifier
- Types via Sorbet
- Fix formula insertion on Excel
- Fix modifier string quoting
- Fix broken Yard doc generation
- Fix: multiple modifiers on the same row weren't being handled
## v0.1.2
- var=... modifier which allows binding a variable to a cell
- Improved error handling and messages
- Moving in a direction that allows for the context-dependent aspects of modifiers
- Fixes a bug with creating a new excel spreadsheet
- Docs & tests
## v0.1.1
- Better support for the various infix operators (+,-,/,*,^,%,=,<,etc)
* Previously we were converting them to their prefix equivalent (multiply, minus, concat, etc)
but excel doesn't support most of those. So we keep them infix
* Didn't support some infix operators (^, %, </>/<=/>=/<>)
* Proper support for operator precedence
- When in verbose mode, print a summary of compiled functions and variables
- docs & tests
## v0.1.0
- revamp of builtin functions
- docs & tests
## v0.0.5
- Support the --backup/-b option
- bin/csvpp (which does the same thing as bin/csv++ but will work better on other filesystems)
- Fix links in gemspec (which end up on rubygems.org)
- docs & tests
## v0.0.4
- Excel support
## v0.0.3
- Fix the gem package to include the bin/ file
## v0.0.2
- First publish to rubygems.org