aboutsummaryrefslogtreecommitdiff
path: root/README.md
diff options
context:
space:
mode:
authorkartofen <kartofen.mail.0@protonmail.com>2025-10-08 20:21:40 +0300
committerkartofen <kartofen.mail.0@protonmail.com>2025-10-08 20:21:40 +0300
commitbcd3391e98511f5913149b87eb8617aaaa4beb6b (patch)
treee21a9be998a2ae9a4007b7c631c64dcb5f9b7277 /README.md
parentadcadb01becb36c79d881080d399328de051e301 (diff)
edit README and build extend ninja.m4
Diffstat (limited to 'README.md')
-rw-r--r--README.md12
1 files changed, 5 insertions, 7 deletions
diff --git a/README.md b/README.md
index 9e853ba..417c644 100644
--- a/README.md
+++ b/README.md
@@ -13,23 +13,21 @@ techniques and ways to add semanitic meaning.
### TODO
-#### Goals
-
-- The LR parser implementation is very dirty and bad
- The building process is too compilated, the grammar parser
should also do the table generation
-- Proper LALR generation
+- Proper LALR generation (without generating all CLR states)
- LL table generation and parsing
- Possibly recursive ascent and recursive descent generation (a bit pointless)
- Left corner / resursive ascent-descent generation (Horspool)
+
- Proper attribute grammar implementation, evaluation, and dealing with cycles
- (S)GLR - Scannerless Generalized LR (Masaru Tomita)
- It would be good to implemented a compiler of C language (C, B, BCPL),
and a high level language, maybe Prolog or something mine and with all
- beingfairly optimized
+ being fairly optimized
### Buildling
-The build script couldn't be more straight forward, use just ```./build.sh```
-and uncomment the lines for the give use
+It is done using Ninja and M4, use `ninja` to build (and run).
+See `ninja.m4` for more information