Without this bugfix, any code that throws the default error, but which haven't
imported the symb...

docs(Theory): Fix NFA state number errors, spelling, and add alt-texts to images

docs(Regex): f...

In BioJulia discussion 11, and BioJulia/Contributing, we've decided to make the
template less ve...

Issue 125 revealed several issues with the existing tokenizer implementation,
several of which a...

Correct automa image link.

Special macros like `@abspos` previously inserted an assert that the buffer's
mark had been set....

I've come to the conclusion that Julia does not make it possible to robustly
check what CPU inst...

Add macros with the same name as pseudomacros.
These return a string explaining that the pseudom...

Instead of buffering an entire line, simply keep track of the number of columns
cleared from the...

The existence of these as real macros does not actually work with arbitrary
`CodeGenContext`s.
T...

Previously, `Machine`s contained the fields `start_state::Int` and
`states::UnitRange{Int}`. How...

This module doesn't make any sense in v1 for two reasons:
First, we can not longer consider the ...

Since Automa doesn't just extend an existing function but adds a new function
based on Transcodi...

Allow preconditions to be set to `:enter` only, and have directly conflicting
preconditions reso...

The user probably wants to use the input error code in most use cases, except
when running a mac...

Before this PR, a user could forget or mis-spell a symbol in the action Dict
passed to generate_...

The `re2nfa` function will create NFAs which include useless eps nodes.
Here, I define a useless...

NFAs with ambiguities often contain multiple ambiguities. Displaying the
simplest ambiguity when...

If two edges in equivalent paths have distinct preconditions, these could be
used to distinguish...

Currently, the functions `re2nfa` and `nfa2dfa` can produce dead (unreachable)
nodes, which is p...

Honestly, four generators and a tokenizer was excessive.
Inline generator was probably not reall...

Many potential users of Automa are not interested in parsing from IOs, but only
buffers. For tho...

An oversight in the ambiguity check meant that actions placed on non-epsilon
edges were accident...

One of the issues with using Automa is that its lack of exports makes it unclear
what is interna...

p_eof only have two valid values: -1 and p_end. By instead using is_eof::Bool
to distinguish the...

Currently, Automa allows you to add a :final action to a regex like "a*", where
it is not possib...

Instead of doing `re.actions[:exit] = [:foo]`, do `onexit(re, :foo)`. It's
cleaner, and avoids m...

nfa2dfa now errors if any RE object has a .actions field with an unsupported
key. This prevents ...

The tokenizer has a completely new design and API.
* It's now much easier to use
* It's now lazy...

With the :goto generator, checking bounds is not permitted anyway.
And the table generator is so...

There is no reason to use `import` over `using`. On the contrary, it muddles
the difference betw...

Users should not have access to the module directly. Instead, export the RE
struct, and also all...

The gensym'd symbols in Automa's constant expressions are computed on
precompilation of Automa.
...

This dependency was used for something completely trivial.

It is possible for users to construct a regex that matches the null set, i.e.
no strings, not ev...

These two data types were originally added to prevent nondeterminism when
compiling regex.
Inste...

The code in generate_reader did not respect the Vars or CodeGenContext.

A current limitation of Automa is that it can be difficult to understand why
a certain string do...

Now prints the minimal possible input to trigger the ambiguity.

I want to release one more non-breaking release before v1, which requires us
to remove the break...

Most generators have < 128 states, basically all have < 32,000. This suggests
we should be able...

Automa's dependency ScanByte implements a ByteSet. Use it instead.

* Disallow unicode in parse_class
* Allow byte ranges in class

* Fix typo geybyte to getbyte
* Use optimized final state check for Tokenizer
* Make Tokenizer...

* Remove unused argument generate_table_code
* Do not needlessly specialize on args in Tokenize...

# Closes #88

In the code generated by generate_table_code for Tokenizer, if the loop exits
due to reaching EO...

Optimize away dict lookup when possible

Fix execute (issue 78)