More work on scene parsing.

Using Nom to help with parsing the contents of leaf nodes.
2016-03-10 23:58:19 -08:00 · 2016-03-10 23:58:19 -08:00 · 38d33ed144
commit 38d33ed144
parent 25f8e0782a
8 changed files with 503 additions and 21 deletions
--- a/Cargo.lock
+++ b/Cargo.lock
@ -3,15 +3,16 @@ name = "psychopath"
 version = "0.1.0"
 dependencies = [
 "docopt 0.6.78 (registry+https://github.com/rust-lang/crates.io-index)",
- "rustc-serialize 0.3.16 (registry+https://github.com/rust-lang/crates.io-index)",
+ "nom 1.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
+ "rustc-serialize 0.3.18 (registry+https://github.com/rust-lang/crates.io-index)",
 ]

 [[package]]
 name = "aho-corasick"
-version = "0.4.0"
+version = "0.5.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
- "memchr 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)",
+ "memchr 0.1.10 (registry+https://github.com/rust-lang/crates.io-index)",
 ]

 [[package]]
@ -19,42 +20,48 @@ name = "docopt"
 version = "0.6.78"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
- "regex 0.1.44 (registry+https://github.com/rust-lang/crates.io-index)",
- "rustc-serialize 0.3.16 (registry+https://github.com/rust-lang/crates.io-index)",
+ "regex 0.1.55 (registry+https://github.com/rust-lang/crates.io-index)",
+ "rustc-serialize 0.3.18 (registry+https://github.com/rust-lang/crates.io-index)",
 "strsim 0.3.0 (registry+https://github.com/rust-lang/crates.io-index)",
 ]

 [[package]]
 name = "libc"
-version = "0.2.4"
+version = "0.2.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"

 [[package]]
 name = "memchr"
-version = "0.1.7"
+version = "0.1.10"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
- "libc 0.2.4 (registry+https://github.com/rust-lang/crates.io-index)",
+ "libc 0.2.8 (registry+https://github.com/rust-lang/crates.io-index)",
 ]

+[[package]]
+name = "nom"
+version = "1.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+
 [[package]]
 name = "regex"
-version = "0.1.44"
+version = "0.1.55"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
- "aho-corasick 0.4.0 (registry+https://github.com/rust-lang/crates.io-index)",
- "memchr 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)",
- "regex-syntax 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
+ "aho-corasick 0.5.1 (registry+https://github.com/rust-lang/crates.io-index)",
+ "memchr 0.1.10 (registry+https://github.com/rust-lang/crates.io-index)",
+ "regex-syntax 0.2.5 (registry+https://github.com/rust-lang/crates.io-index)",
+ "utf8-ranges 0.1.3 (registry+https://github.com/rust-lang/crates.io-index)",
 ]

 [[package]]
 name = "regex-syntax"
-version = "0.2.2"
+version = "0.2.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"

 [[package]]
 name = "rustc-serialize"
-version = "0.3.16"
+version = "0.3.18"
 source = "registry+https://github.com/rust-lang/crates.io-index"

 [[package]]
@ -62,3 +69,8 @@ name = "strsim"
 version = "0.3.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"

+[[package]]
+name = "utf8-ranges"
+version = "0.1.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+
--- a/Cargo.toml
+++ b/Cargo.toml
@ -5,4 +5,5 @@ authors = ["Nathan Vegdahl <cessen@cessen.com>"]

 [dependencies]
 docopt = "0.6"
-rustc-serialize = "0.3"
+rustc-serialize = "0.3"
+nom = "1.2"
--- a/src/main.rs
+++ b/src/main.rs
@ -1,5 +1,7 @@
 extern crate rustc_serialize;
 extern crate docopt;
+#[macro_use]
+extern crate nom;

 mod math;
 mod algorithm;
--- a/src/parse/basics.rs
+++ b/src/parse/basics.rs
@ -0,0 +1,289 @@
+//! Some basic nom parsers
+#![allow(dead_code)]
+
+use std::str;
+
+use nom::{IResult, Needed, digit, multispace};
+use nom::IResult::*;
+
+
+// Parsers for numbers surrounded by whitespace
+named!(pub ws_u32<u32>, delimited!(opt!(multispace), u32_utf8, opt!(multispace)));
+named!(pub ws_u64<u64>, delimited!(opt!(multispace), u64_utf8, opt!(multispace)));
+named!(pub ws_i32<i32>, delimited!(opt!(multispace), i32_utf8, opt!(multispace)));
+named!(pub ws_i64<i64>, delimited!(opt!(multispace), i64_utf8, opt!(multispace)));
+named!(pub ws_f32<f32>, delimited!(opt!(multispace), f32_utf8, opt!(multispace)));
+named!(pub ws_f64<f64>, delimited!(opt!(multispace), f64_utf8, opt!(multispace)));
+
+
+
+
+// ========================================================
+
+named!(pub u32_utf8<u32>, chain!(
+        bytes: digit,
+        || { str::from_utf8(bytes).unwrap().parse::<u32>().unwrap() }
+));
+
+named!(pub i32_utf8<i32>, chain!(
+        sign: one_of!("-+")? ~
+        bytes: digit,
+        || {
+            match sign {
+                Some(s) if s == '-' => -str::from_utf8(bytes).unwrap().parse::<i32>().unwrap(),
+                _ => str::from_utf8(bytes).unwrap().parse::<i32>().unwrap(),
+            }
+        }
+));
+
+named!(pub u64_utf8<u64>, chain!(
+        bytes: digit,
+        || { str::from_utf8(bytes).unwrap().parse::<u64>().unwrap() }
+));
+
+named!(pub i64_utf8<i64>, chain!(
+        sign: one_of!("-+")? ~
+        bytes: digit,
+        || {
+            match sign {
+                Some(s) if s == '-' => -str::from_utf8(bytes).unwrap().parse::<i64>().unwrap(),
+                _ => str::from_utf8(bytes).unwrap().parse::<i64>().unwrap(),
+            }
+        }
+));
+
+named!(pub f32_utf8<f32>, chain!(
+        bytes: take_decimal_real,
+        || {
+            str::from_utf8(bytes).unwrap().parse::<f32>().unwrap()
+        }
+));
+
+named!(pub f64_utf8<f64>, chain!(
+        bytes: take_decimal_real,
+        || {
+            str::from_utf8(bytes).unwrap().parse::<f64>().unwrap()
+        }
+));
+
+fn take_decimal_integer(i: &[u8]) -> IResult<&[u8], &[u8]> {
+    named!(rr<&[u8], ()>, chain!(
+        one_of!("-+")? ~
+        digit,
+        ||{()}
+    ));
+
+    match rr(i) {
+        Done(remaining, _) => {
+            let m = i.len() - remaining.len();
+            if m == 0 {
+                Incomplete(Needed::Unknown)
+            } else {
+                Done(&i[m..], &i[0..m])
+            }
+        }
+
+        Error(e) => Error(e),
+
+        Incomplete(e) => Incomplete(e),
+    }
+}
+
+fn take_decimal_real(i: &[u8]) -> IResult<&[u8], &[u8]> {
+    named!(rr<&[u8], ()>, chain!(
+        one_of!("-+")? ~
+        digit ~
+        complete!(chain!(tag!(".") ~ digit, ||{()}))?,
+        ||{()}
+    ));
+
+    match rr(i) {
+        Done(remaining, _) => {
+            let m = i.len() - remaining.len();
+            if m == 0 {
+                Incomplete(Needed::Unknown)
+            } else {
+                Done(&i[m..], &i[0..m])
+            }
+        }
+
+        Error(e) => Error(e),
+
+        Incomplete(e) => Incomplete(e),
+    }
+}
+
+
+
+
+// ========================================================
+
+#[cfg(test)]
+mod test {
+    use nom::IResult;
+    use nom::IResult::*;
+    use super::take_decimal_real;
+    use super::*;
+
+    #[test]
+    fn ws_u32_1() {
+        assert_eq!(ws_u32(b"42"), Done(&b""[..], 42));
+        assert_eq!(ws_u32(b"     42"), Done(&b""[..], 42));
+        assert_eq!(ws_u32(b"42   "), Done(&b""[..], 42));
+        assert_eq!(ws_u32(b"     42"), Done(&b""[..], 42));
+        assert_eq!(ws_u32(b"     42   53"), Done(&b"53"[..], 42));
+    }
+
+    #[test]
+    fn ws_i32_1() {
+        assert_eq!(ws_i32(b"42"), Done(&b""[..], 42));
+        assert_eq!(ws_i32(b"     42"), Done(&b""[..], 42));
+        assert_eq!(ws_i32(b"42   "), Done(&b""[..], 42));
+        assert_eq!(ws_i32(b"     42"), Done(&b""[..], 42));
+        assert_eq!(ws_i32(b"     42   53"), Done(&b"53"[..], 42));
+    }
+
+    #[test]
+    fn ws_i32_2() {
+        assert_eq!(ws_i32(b"-42"), Done(&b""[..], -42));
+        assert_eq!(ws_i32(b"     -42"), Done(&b""[..], -42));
+        assert_eq!(ws_i32(b"-42   "), Done(&b""[..], -42));
+        assert_eq!(ws_i32(b"     -42   "), Done(&b""[..], -42));
+        assert_eq!(ws_i32(b"     -42   53"), Done(&b"53"[..], -42));
+    }
+
+    #[test]
+    fn ws_u64_1() {
+        assert_eq!(ws_u64(b"42"), Done(&b""[..], 42));
+        assert_eq!(ws_u64(b"     42"), Done(&b""[..], 42));
+        assert_eq!(ws_u64(b"42   "), Done(&b""[..], 42));
+        assert_eq!(ws_u64(b"     42"), Done(&b""[..], 42));
+        assert_eq!(ws_u64(b"     42   53"), Done(&b"53"[..], 42));
+    }
+
+    #[test]
+    fn ws_i64_1() {
+        assert_eq!(ws_i64(b"42"), Done(&b""[..], 42));
+        assert_eq!(ws_i64(b"     42"), Done(&b""[..], 42));
+        assert_eq!(ws_i64(b"42   "), Done(&b""[..], 42));
+        assert_eq!(ws_i64(b"     42"), Done(&b""[..], 42));
+        assert_eq!(ws_i64(b"     42   53"), Done(&b"53"[..], 42));
+    }
+
+    #[test]
+    fn ws_i64_2() {
+        assert_eq!(ws_i64(b"-42"), Done(&b""[..], -42));
+        assert_eq!(ws_i64(b"     -42"), Done(&b""[..], -42));
+        assert_eq!(ws_i64(b"-42   "), Done(&b""[..], -42));
+        assert_eq!(ws_i64(b"     -42   "), Done(&b""[..], -42));
+        assert_eq!(ws_i64(b"     -42   53"), Done(&b"53"[..], -42));
+    }
+
+    #[test]
+    fn ws_f32_1() {
+        assert_eq!(ws_f32(b"42"), Done(&b""[..], 42.0));
+        assert_eq!(ws_f32(b"     42"), Done(&b""[..], 42.0));
+        assert_eq!(ws_f32(b"42   "), Done(&b""[..], 42.0));
+        assert_eq!(ws_f32(b"     42"), Done(&b""[..], 42.0));
+        assert_eq!(ws_f32(b"     42   53"), Done(&b"53"[..], 42.0));
+    }
+
+    #[test]
+    fn ws_f32_2() {
+        assert_eq!(ws_f32(b"42.5"), Done(&b""[..], 42.5));
+        assert_eq!(ws_f32(b"     42.5"), Done(&b""[..], 42.5));
+        assert_eq!(ws_f32(b"42.5   "), Done(&b""[..], 42.5));
+        assert_eq!(ws_f32(b"     42.5"), Done(&b""[..], 42.5));
+        assert_eq!(ws_f32(b"     42.5   53"), Done(&b"53"[..], 42.5));
+    }
+
+    #[test]
+    fn ws_f32_3() {
+        assert_eq!(ws_f32(b"-42.5"), Done(&b""[..], -42.5));
+        assert_eq!(ws_f32(b"     -42.5"), Done(&b""[..], -42.5));
+        assert_eq!(ws_f32(b"-42.5   "), Done(&b""[..], -42.5));
+        assert_eq!(ws_f32(b"     -42.5"), Done(&b""[..], -42.5));
+        assert_eq!(ws_f32(b"     -42.5   53"), Done(&b"53"[..], -42.5));
+    }
+
+    #[test]
+    fn ws_f64_1() {
+        assert_eq!(ws_f64(b"42"), Done(&b""[..], 42.0));
+        assert_eq!(ws_f64(b"     42"), Done(&b""[..], 42.0));
+        assert_eq!(ws_f64(b"42   "), Done(&b""[..], 42.0));
+        assert_eq!(ws_f64(b"     42"), Done(&b""[..], 42.0));
+        assert_eq!(ws_f64(b"     42   53"), Done(&b"53"[..], 42.0));
+    }
+
+    #[test]
+    fn ws_f64_2() {
+        assert_eq!(ws_f64(b"42.5"), Done(&b""[..], 42.5));
+        assert_eq!(ws_f64(b"     42.5"), Done(&b""[..], 42.5));
+        assert_eq!(ws_f64(b"42.5   "), Done(&b""[..], 42.5));
+        assert_eq!(ws_f64(b"     42.5"), Done(&b""[..], 42.5));
+        assert_eq!(ws_f64(b"     42.5   53"), Done(&b"53"[..], 42.5));
+    }
+
+    #[test]
+    fn ws_f64_3() {
+        assert_eq!(ws_f64(b"-42.5"), Done(&b""[..], -42.5));
+        assert_eq!(ws_f64(b"     -42.5"), Done(&b""[..], -42.5));
+        assert_eq!(ws_f64(b"-42.5   "), Done(&b""[..], -42.5));
+        assert_eq!(ws_f64(b"     -42.5"), Done(&b""[..], -42.5));
+        assert_eq!(ws_f64(b"     -42.5   53"), Done(&b"53"[..], -42.5));
+    }
+
+    #[test]
+    fn take_decimal_real_1() {
+        assert_eq!(take_decimal_real(b"-42.3"), Done(&b""[..], &b"-42.3"[..]));
+        assert_eq!(take_decimal_real(b"42.3"), Done(&b""[..], &b"42.3"[..]));
+        assert_eq!(take_decimal_real(b"-42"), Done(&b""[..], &b"-42"[..]));
+        assert_eq!(take_decimal_real(b"+42.3"), Done(&b""[..], &b"+42.3"[..]));
+    }
+
+    #[test]
+    fn u32_utf8_1() {
+        assert_eq!(u32_utf8(b"42"), Done(&b""[..], 42));
+        assert_eq!(u32_utf8(b"-42").is_err(), true);
+    }
+
+    #[test]
+    fn i32_utf8_1() {
+        assert_eq!(i32_utf8(b"42"), Done(&b""[..], 42));
+        assert_eq!(i32_utf8(b"-42"), Done(&b""[..], -42));
+        assert_eq!(i32_utf8(b"+42"), Done(&b""[..], 42));
+        assert_eq!(i32_utf8(b"--42").is_err(), true);
+        assert_eq!(i32_utf8(b"+-42").is_err(), true);
+    }
+
+    #[test]
+    fn u64_utf8_1() {
+        assert_eq!(u64_utf8(b"42"), Done(&b""[..], 42));
+        assert_eq!(u64_utf8(b"-42").is_err(), true);
+    }
+
+    #[test]
+    fn i64_utf8_1() {
+        assert_eq!(i64_utf8(b"42"), Done(&b""[..], 42));
+        assert_eq!(i64_utf8(b"-42"), Done(&b""[..], -42));
+        assert_eq!(i64_utf8(b"+42"), Done(&b""[..], 42));
+        assert_eq!(i64_utf8(b"--42").is_err(), true);
+        assert_eq!(i64_utf8(b"+-42").is_err(), true);
+    }
+
+    #[test]
+    fn f32_utf8_1() {
+        assert_eq!(f32_utf8(b"-42.3"), Done(&b""[..], -42.3));
+        assert_eq!(f32_utf8(b"+42.3"), Done(&b""[..], 42.3));
+        assert_eq!(f32_utf8(b"42.3"), Done(&b""[..], 42.3));
+        assert_eq!(f32_utf8(b"42"), Done(&b""[..], 42.0));
+    }
+
+    #[test]
+    fn f64_utf8_1() {
+        assert_eq!(f64_utf8(b"-42.3"), Done(&b""[..], -42.3));
+        assert_eq!(f64_utf8(b"+42.3"), Done(&b""[..], 42.3));
+        assert_eq!(f64_utf8(b"42.3"), Done(&b""[..], 42.3));
+        assert_eq!(f64_utf8(b"42"), Done(&b""[..], 42.0));
+    }
+}
--- a/src/parse/data_tree.rs
+++ b/src/parse/data_tree.rs
@ -42,15 +42,62 @@ impl<'a> DataTree<'a> {
        }
    }

+    pub fn get_first_child_with_type_name(&'a self, type_name: &str) -> Option<&'a DataTree> {
+        if let &DataTree::Internal { ref children, .. } = self {
+            for child in children.iter() {
+                match child {
+                    &DataTree::Internal { type_name: tn, .. } => {
+                        if tn == type_name {
+                            return Some(child);
+                        }
+                    }
+
+                    &DataTree::Leaf { type_name: tn, .. } => {
+                        if tn == type_name {
+                            return Some(child);
+                        }
+                    }
+                }
+            }
+            return None;
+        } else {
+            return None;
+        }
+    }
+
+    pub fn count_children_with_type_name(&self, type_name: &str) -> usize {
+        if let &DataTree::Internal { ref children, .. } = self {
+            let mut count = 0;
+            for child in children.iter() {
+                match child {
+                    &DataTree::Internal { type_name: tn, .. } => {
+                        if tn == type_name {
+                            count += 1;
+                        }
+                    }
+
+                    &DataTree::Leaf { type_name: tn, .. } => {
+                        if tn == type_name {
+                            count += 1;
+                        }
+                    }
+                }
+            }
+            return count;
+        } else {
+            return 0;
+        }
+    }
+
    // For unit tests
-    fn internal_data(&'a self) -> (&'a str, Option<&'a str>, &'a Vec<DataTree<'a>>) {
+    fn internal_data_or_panic(&'a self) -> (&'a str, Option<&'a str>, &'a Vec<DataTree<'a>>) {
        if let DataTree::Internal { type_name, ident, ref children } = *self {
            (type_name, ident, children)
        } else {
            panic!("Expected DataTree::Internal, found DataTree::Leaf")
        }
    }
-    fn leaf_data(&'a self) -> (&'a str, &'a str) {
+    fn leaf_data_or_panic(&'a self) -> (&'a str, &'a str) {
        if let DataTree::Leaf { type_name, contents } = *self {
            (type_name, contents)
        } else {
@ -451,13 +498,13 @@ mod tests {
        let dt = DataTree::from_str(input).unwrap();

        // Root
-        let (t, i, c) = dt.internal_data();
+        let (t, i, c) = dt.internal_data_or_panic();
        assert_eq!(t, "ROOT");
        assert_eq!(i, None);
        assert_eq!(c.len(), 1);

        // First (and only) child
-        let (t, i, c) = c[0].internal_data();
+        let (t, i, c) = c[0].internal_data_or_panic();
        assert_eq!(t, "Thing");
        assert_eq!(i, None);
        assert_eq!(c.len(), 0);
--- a/src/parse/mod.rs
+++ b/src/parse/mod.rs
@ -1,4 +1,5 @@
 mod data_tree;
 mod psy;
+pub mod basics;

 pub use self::data_tree::DataTree;
--- a/src/parse/psy.rs
+++ b/src/parse/psy.rs
@ -1,8 +1,137 @@
 #![allow(dead_code)]

+use std::result::Result;
+
+use nom;
+use nom::IResult;
+
 use renderer::Renderer;
 use super::DataTree;
+use super::basics::ws_u32;

-pub fn parse_frame(tree: &DataTree) -> Renderer {
-    unimplemented!()
+
+/// Takes in a DataTree representing a Scene node and returns
+/// a renderer.
+pub fn parse_frame(tree: &DataTree) -> Result<Renderer, ()> {
+    // Verify we have the right number of each section
+    if tree.count_children_with_type_name("Output") != 1 {
+        return Err(());
+    }
+    if tree.count_children_with_type_name("RenderSettings") != 1 {
+        return Err(());
+    }
+    if tree.count_children_with_type_name("Camera") != 1 {
+        return Err(());
+    }
+    if tree.count_children_with_type_name("World") != 1 {
+        return Err(());
+    }
+    if tree.count_children_with_type_name("Assembly") != 1 {
+        return Err(());
+    }
+
+    // Parse output info
+    let output_info = if let &DataTree::Internal{ref children, ..} =
+                             tree.get_first_child_with_type_name("Output").unwrap() {
+        let mut found_path = false;
+        let mut path = String::new();
+
+        for child in children {
+            match child {
+                &DataTree::Leaf { type_name, contents } if type_name == "Path" => {
+                    // TODO: proper string escaping and quotes stripping
+                    found_path = true;
+                    path = contents.to_string();
+                }
+
+                &DataTree::Leaf { type_name, contents } if type_name == "FileFormat" => {
+                    // TODO
+                    unimplemented!()
+                }
+
+                &DataTree::Leaf { type_name, contents } if type_name == "ColorSpace" => {
+                    // TODO
+                    unimplemented!()
+                }
+
+                &DataTree::Leaf { type_name, contents } if type_name == "Dither" => {
+                    // TODO
+                    unimplemented!()
+                }
+
+                _ => {}
+            }
+        }
+
+        if found_path {
+            (path)
+        } else {
+            return Err(());
+        }
+    } else {
+        return Err(());
+    };
+
+    // Parse render settings
+    let render_settings = if let &DataTree::Internal{ref children, ..} =
+                                 tree.get_first_child_with_type_name("RenderSettings").unwrap() {
+        let mut found_res = false;
+        let mut found_spp = false;
+        let mut res = (0, 0);
+        let mut spp = 0;
+        let mut seed = 0;
+
+        for child in children {
+            match child {
+                // Resolution
+                &DataTree::Leaf { type_name, contents } if type_name == "Resolution" => {
+                    if let IResult::Done(_, (w, h)) = closure!(terminated!(tuple!(ws_u32, ws_u32),
+                                                nom::eof))(contents.as_bytes()) {
+                        found_res = true;
+                        res = (w, h);
+                    } else {
+                        // Found Resolution, but its contents is not in the right format
+                        return Err(());
+                    }
+                }
+
+                &DataTree::Leaf { type_name, contents } if type_name == "SamplesPerPixel" => {
+                    if let IResult::Done(_, n) = ws_u32(contents.as_bytes()) {
+                        found_spp = true;
+                        spp = n;
+                    } else {
+                        // Found SamplesPerPixel, but its contents is not in the right format
+                        return Err(());
+                    }
+                }
+
+                &DataTree::Leaf { type_name, contents } if type_name == "Seed" => {
+                    if let IResult::Done(_, n) = ws_u32(contents.as_bytes()) {
+                        seed = n;
+                    } else {
+                        // Found Seed, but its contents is not in the right format
+                        return Err(());
+                    }
+                }
+
+                _ => {}
+            }
+        }
+
+        if found_res && found_spp {
+            (res, spp)
+        } else {
+            return Err(());
+        }
+    } else {
+        return Err(());
+    };
+
+    // Parse camera
+
+    // Parse world
+
+    // Parse root scene assembly
+
+    return Err(());
 }
--- a/src/renderer.rs
+++ b/src/renderer.rs
@ -8,6 +8,7 @@ use math::fast_logit;
 use image::Image;
 use surface;
 use surface::Surface;
+use scene::Scene;

 pub struct Renderer {
    pub output_file: String,