Publish version 0.1.5.

This test will grab documents from external sources and compare Organic's parser vs the official org-mode parser to ensure they are parsing the same. This is so we do not introduce large irrelevant documents in the git history and so we do not introduce documents with restrictive licenses into the repository.

.lighthouse/pipeline-foreign-document-test.yaml

@@ -0,0 +1,203 @@
+apiVersion: tekton.dev/v1beta1
+kind: PipelineRun
+metadata:
+  name: rust-foreign-document-test
+spec:
+  pipelineSpec:
+    timeouts:
+      pipeline: "2h0m0s"
+      tasks: "1h0m40s"
+      finally: "0h30m0s"
+    params:
+      - name: image-name
+        description: The name for the built image
+        type: string
+      - name: path-to-image-context
+        description: The path to the build context
+        type: string
+      - name: path-to-dockerfile
+        description: The path to the Dockerfile
+        type: string
+    tasks:
+      - name: do-stuff
+        taskSpec:
+          metadata: {}
+          stepTemplate:
+            image: alpine:3.18
+            name: ""
+            resources:
+              requests:
+                cpu: 10m
+                memory: 600Mi
+            workingDir: /workspace/source
+          steps:
+            - image: alpine:3.18
+              name: do-stuff-step
+              script: |
+                #!/usr/bin/env sh
+                echo "hello world"
+      - name: report-pending
+        taskRef:
+          name: gitea-set-status
+        runAfter:
+          - fetch-repository
+        params:
+          - name: CONTEXT
+            value: "$(params.JOB_NAME)"
+          - name: REPO_FULL_NAME
+            value: "$(params.REPO_OWNER)/$(params.REPO_NAME)"
+          - name: GITEA_HOST_URL
+            value: code.fizz.buzz
+          - name: SHA
+            value: "$(tasks.fetch-repository.results.commit)"
+          - name: DESCRIPTION
+            value: "Build $(params.JOB_NAME) has started"
+          - name: STATE
+            value: pending
+          - name: TARGET_URL
+            value: "https://tekton.fizz.buzz/#/namespaces/$(context.pipelineRun.namespace)/pipelineruns/$(context.pipelineRun.name)"
+      - name: fetch-repository
+        taskRef:
+          name: git-clone
+        workspaces:
+          - name: output
+            workspace: git-source
+        params:
+          - name: url
+            value: $(params.REPO_URL)
+          - name: revision
+            value: $(params.PULL_BASE_SHA)
+          - name: deleteExisting
+            value: "true"
+      - name: build-image
+        taskRef:
+          name: kaniko
+        params:
+          - name: IMAGE
+            value: "$(params.image-name):$(tasks.fetch-repository.results.commit)"
+          - name: CONTEXT
+            value: $(params.path-to-image-context)
+          - name: DOCKERFILE
+            value: $(params.path-to-dockerfile)
+          - name: BUILDER_IMAGE
+            value: "gcr.io/kaniko-project/executor:v1.12.1"
+          - name: EXTRA_ARGS
+            value:
+              - --target=foreign-document-test
+              - --cache=true
+              - --cache-copy-layers
+              - --cache-repo=harbor.fizz.buzz/kanikocache/cache
+              - --use-new-run # Should result in a speed-up
+              - --reproducible # To remove timestamps so layer caching works.
+              - --snapshot-mode=redo
+              - --skip-unused-stages=true
+              - --registry-mirror=dockerhub.dockerhub.svc.cluster.local
+        workspaces:
+          - name: source
+            workspace: git-source
+          - name: dockerconfig
+            workspace: docker-credentials
+        runAfter:
+          - fetch-repository
+      - name: run-image
+        taskRef:
+          name: run-docker-image
+        workspaces:
+          - name: source
+            workspace: git-source
+          - name: cargo-cache
+            workspace: cargo-cache
+        runAfter:
+          - build-image
+        params:
+          - name: docker-image
+            value: "$(params.image-name):$(tasks.fetch-repository.results.commit)"
+    finally:
+      - name: report-success
+        when:
+          - input: "$(tasks.status)"
+            operator: in
+            values: ["Succeeded", "Completed"]
+        taskRef:
+          name: gitea-set-status
+        params:
+          - name: CONTEXT
+            value: "$(params.JOB_NAME)"
+          - name: REPO_FULL_NAME
+            value: "$(params.REPO_OWNER)/$(params.REPO_NAME)"
+          - name: GITEA_HOST_URL
+            value: code.fizz.buzz
+          - name: SHA
+            value: "$(tasks.fetch-repository.results.commit)"
+          - name: DESCRIPTION
+            value: "Build $(params.JOB_NAME) has succeeded"
+          - name: STATE
+            value: success
+          - name: TARGET_URL
+            value: "https://tekton.fizz.buzz/#/namespaces/$(context.pipelineRun.namespace)/pipelineruns/$(context.pipelineRun.name)"
+      - name: report-failure
+        when:
+          - input: "$(tasks.status)"
+            operator: in
+            values: ["Failed"]
+        taskRef:
+          name: gitea-set-status
+        params:
+          - name: CONTEXT
+            value: "$(params.JOB_NAME)"
+          - name: REPO_FULL_NAME
+            value: "$(params.REPO_OWNER)/$(params.REPO_NAME)"
+          - name: GITEA_HOST_URL
+            value: code.fizz.buzz
+          - name: SHA
+            value: "$(tasks.fetch-repository.results.commit)"
+          - name: DESCRIPTION
+            value: "Build $(params.JOB_NAME) has failed"
+          - name: STATE
+            value: failure
+          - name: TARGET_URL
+            value: "https://tekton.fizz.buzz/#/namespaces/$(context.pipelineRun.namespace)/pipelineruns/$(context.pipelineRun.name)"
+      - name: cargo-cache-autoclean
+        taskRef:
+          name: run-docker-image
+        workspaces:
+          - name: source
+            workspace: git-source
+          - name: cargo-cache
+            workspace: cargo-cache
+        params:
+          - name: command
+            value: [cargo, cache, --autoclean]
+          - name: args
+            value: []
+          - name: docker-image
+            value: "$(params.image-name):$(tasks.fetch-repository.results.commit)"
+    workspaces:
+      - name: git-source
+      - name: docker-credentials
+      - name: cargo-cache
+  workspaces:
+    - name: git-source
+      volumeClaimTemplate:
+        spec:
+          storageClassName: "nfs-client"
+          accessModes:
+            - ReadWriteOnce
+          resources:
+            requests:
+              storage: 10Gi
+      subPath: rust-source
+    - name: cargo-cache
+      persistentVolumeClaim:
+        claimName: organic-cargo-cache-test-foreign-document
+    - name: docker-credentials
+      secret:
+        secretName: harbor-plain
+  serviceAccountName: build-bot
+  params:
+    - name: image-name
+      value: "harbor.fizz.buzz/private/organic-test-foreign-document"
+    - name: path-to-image-context
+      value: docker/organic_test/
+    - name: path-to-dockerfile
+      value: docker/organic_test/Dockerfile

.lighthouse/pipeline-rust-test.yaml

@@ -83,6 +83,7 @@ spec:
             value: "gcr.io/kaniko-project/executor:v1.12.1"
           - name: EXTRA_ARGS
             value:
+              - --target=tester
               - --cache=true
               - --cache-copy-layers
               - --cache-repo=harbor.fizz.buzz/kanikocache/cache

.lighthouse/triggers.yaml

@@ -16,6 +16,13 @@ spec:
       skip_branches:
         # We already run on every commit, so running when the semver tags get pushed is causing needless double-processing.
         - "^v[0-9]+\\.[0-9]+\\.[0-9]+$"
+    - name: rust-foreign-document-test
+      source: "pipeline-foreign-document-test.yaml"
+      # Override https-based url from lighthouse events.
+      clone_uri: "git@code.fizz.buzz:talexander/organic.git"
+      skip_branches:
+        # We already run on every commit, so running when the semver tags get pushed is causing needless double-processing.
+        - "^v[0-9]+\\.[0-9]+\\.[0-9]+$"
     - name: rust-build
       source: "pipeline-rust-build.yaml"
       # Override https-based url from lighthouse events.

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "organic"
-version = "0.1.4"
+version = "0.1.5"
 authors = ["Tom Alexander <tom@fizz.buzz>"]
 description = "An org-mode parser."
 edition = "2021"

Makefile

@@ -42,6 +42,10 @@ dockertest:
 > $(MAKE) -C docker/organic_test
 > docker run --init --rm -i -t --read-only -v "$$(readlink -f ./):/source:ro" --mount type=tmpfs,destination=/tmp --mount source=cargo-cache,target=/usr/local/cargo/registry --mount source=rust-cache,target=/target --env CARGO_TARGET_DIR=/target -w /source organic-test --no-default-features --features compare --no-fail-fast --lib --test test_loader -- --test-threads $(TESTJOBS)
 
+.PHONY: foreign_document_test
+foreign_document_test:
+> $(MAKE) -C docker/organic_test run_foreign_document_test
+
 .PHONY: dockerclean
 dockerclean:
 # Delete volumes created for running the tests in docker. This does not touch anything related to the jaeger docker container.

docker/cargo_fmt/Makefile

@@ -6,7 +6,7 @@ all: build push
 
 .PHONY: build
 build:
-	docker build -t $(IMAGE_NAME) -f Dockerfile ../../
+	docker build -t $(IMAGE_NAME) -f Dockerfile .
 
 .PHONY: push
 push:

docker/organic_build/Makefile

@@ -6,7 +6,7 @@ all: build push
 
 .PHONY: build
 build:
-	docker build -t $(IMAGE_NAME) -f Dockerfile ../../
+	docker build -t $(IMAGE_NAME) -f Dockerfile .
 
 .PHONY: push
 push:

docker/organic_test/Dockerfile

@@ -25,7 +25,7 @@ RUN make compile
 RUN make DESTDIR="/root/dist" install
 
 
-FROM rustlang/rust:nightly-alpine3.17
+FROM rustlang/rust:nightly-alpine3.17 AS tester
 ENV LANG=en_US.UTF-8
 RUN apk add --no-cache musl-dev ncurses gnutls
 RUN cargo install --locked --no-default-features --features ci-autoclean cargo-cache
@@ -33,3 +33,16 @@ COPY --from=build-emacs /root/dist/ /
 COPY --from=build-org-mode /root/dist/ /
 
 ENTRYPOINT ["cargo", "test"]
+
+
+FROM build as foreign-document-gather
+RUN mkdir /foreign_documents
+
+
+FROM tester as foreign-document-test
+RUN apk add --no-cache bash
+RUN mkdir /foreign_documents
+COPY --from=build-org-mode /root/org-mode/doc /foreign_documents/org-mode
+COPY foreign_document_test_entrypoint.sh /entrypoint.sh
+RUN chmod +x /entrypoint.sh
+ENTRYPOINT ["/entrypoint.sh"]

docker/organic_test/Makefile

@@ -6,7 +6,11 @@ all: build push
 
 .PHONY: build
 build:
-	docker build -t $(IMAGE_NAME) -f Dockerfile ../../
+	docker build -t $(IMAGE_NAME) -f Dockerfile --target tester .
+
+.PHONY: build_foreign_document_test
+build_foreign_document_test:
+	docker build -t $(IMAGE_NAME)-foreign-document -f Dockerfile --target foreign-document-test .
 
 .PHONY: push
 push:
@@ -34,3 +38,7 @@ run: build
 .PHONY: shell
 shell: build
 	docker run --rm -i -t --entrypoint /bin/sh --mount type=tmpfs,destination=/tmp -v "$$(readlink -f ../../):/source:ro" --workdir=/source --mount source=cargo-cache,target=/usr/local/cargo/registry --mount source=rust-cache,target=/target --env CARGO_TARGET_DIR=/target $(IMAGE_NAME)
+
+.PHONY: run_foreign_document_test
+run_foreign_document_test: build_foreign_document_test
+	docker run --rm --init --read-only --mount type=tmpfs,destination=/tmp -v "$$(readlink -f ../../):/source:ro" --workdir=/source --mount source=cargo-cache,target=/usr/local/cargo/registry --mount source=rust-cache,target=/target --env CARGO_TARGET_DIR=/target $(IMAGE_NAME)-foreign-document

docker/organic_test/foreign_document_test_entrypoint.sh

@@ -0,0 +1,57 @@
+#!/usr/bin/env bash
+#
+# Run the Organic compare script against a series of documents sourced from exterior places.
+set -euo pipefail
+IFS=$'\n\t'
+DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
+
+function log {
+    (>&2 echo "${@}")
+}
+
+function die {
+    local status_code="$1"
+    shift
+    (>&2 echo "${@}")
+    exit "$status_code"
+}
+
+function main {
+    cargo build --no-default-features --features compare --profile release-lto
+    if [ "${CARGO_TARGET_DIR:-}" = "" ]; then
+        CARGO_TARGET_DIR=$(realpath target/)
+    fi
+    PARSE="${CARGO_TARGET_DIR}/release-lto/parse"
+
+    run_compare "org-mode/org-guide.org" "/foreign_documents/org-mode/org-guide.org"
+    run_compare "org-mode/org-manual.org" "/foreign_documents/org-mode/org-manual.org"
+}
+
+function run_compare {
+    local name="$1"
+    local target_document="$2"
+    set +e
+    $PARSE "$target_document" &> /dev/null
+    local status=$?
+    set -e
+    if [ "$status" -eq 0 ]; then
+        echo "$(green_text "GOOD")  $name"
+    else
+        echo "$(red_text "FAIL")  $name"
+        return 1
+    fi
+}
+
+function green_text {
+    (IFS=' '; printf '\x1b[38;2;0;255;0m%s\x1b[0m' "${*}")
+}
+
+function red_text {
+    (IFS=' '; printf '\x1b[38;2;255;0;0m%s\x1b[0m' "${*}")
+}
+
+function yellow_text {
+    (IFS=' '; printf '\x1b[38;2;255;255;0m%s\x1b[0m' "${*}")
+}
+
+main "${@}"

org_mode_samples/README.org

@@ -0,0 +1 @@
+This folder contains org-mode documents that get automatically included as tests using build.rs.

org_mode_samples/object/plain_link/with_trailing_whitespace.org

@@ -0,0 +1 @@
+mailto:foo@bar.baz      .

org_mode_samples/object/text_markup/verbatim_with_equals_inside.org

@@ -1 +1,4 @@
 foo ==>bar=.
+
+# This uses a zero-width space to escape the equals signs to make the verbatim not end.
+=lorem =​= ipsum=

scripts/run_docker_compare_bisect.bash

@@ -0,0 +1,93 @@
+#!/usr/bin/env bash
+#
+# Bisect parsing a file at various line cut-off points to see which line causes the parse to differ from emacs.
+set -euo pipefail
+IFS=$'\n\t'
+DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
+
+cd "$DIR/../"
+REALPATH=$(command -v uu-realpath || command -v realpath)
+
+############## Setup #########################
+
+function cleanup {
+    for f in "${folders[@]}"; do
+        log "Deleting $f"
+        rm -rf "$f"
+    done
+}
+folders=()
+for sig in EXIT INT QUIT HUP TERM; do
+  trap "set +e; cleanup" "$sig"
+done
+
+function die {
+    local status_code="$1"
+    shift
+    (>&2 echo "${@}")
+    exit "$status_code"
+}
+
+function log {
+    (>&2 echo "${@}")
+}
+
+############## Program #########################
+
+function main {
+    log "Is is recommended that the output of \`mktemp -d -t 'compare_bisect.XXXXXXXX'\` is inside a tmpfs filesystem since this script will make many writes to these folders."
+
+    local target_full_path=$($REALPATH "$1")
+    SOURCE_FOLDER=$(dirname "$target_full_path")
+    TARGET_DOCUMENT=$(basename "$target_full_path")
+
+
+    local good=0
+    local bad=$(wc -l "$SOURCE_FOLDER/$TARGET_DOCUMENT" | awk '{print $1}')
+
+    set +e
+    run_parse "$bad" &> /dev/null
+    local status=$?
+    set -e
+    if [ $status -eq 0 ]; then
+        log "Entire file passes."
+        exit 0
+    fi
+
+    while [[ "$((bad - good))" -gt 1 ]]; do
+        local next_line=$((((bad - good) / 2) + good))
+        log "Testing line $next_line"
+        set +e
+        run_parse "$next_line" &> /dev/null
+        local status=$?
+        set -e
+        if [ $status -eq 0 ]; then
+            good="$next_line"
+            log "Line $next_line good"
+        else
+            bad="$next_line"
+            log "Line $next_line bad"
+        fi
+    done
+    echo "Bad line: $bad"
+}
+
+function setup_temp_dir {
+    local temp_dir=$(mktemp -d -t 'compare_bisect.XXXXXXXX')
+    cp -r "$SOURCE_FOLDER/"* "$temp_dir/"
+    echo "$temp_dir"
+}
+
+function run_parse {
+    local lines="$1"
+    local temp_dir=$(setup_temp_dir)
+    folders+=("$temp_dir")
+    cat "$SOURCE_FOLDER/$TARGET_DOCUMENT" | head -n "$lines" > "$temp_dir/$TARGET_DOCUMENT"
+    "${DIR}/run_docker_compare.bash" "$temp_dir/$TARGET_DOCUMENT"
+    local status=$?
+    rm -rf "$temp_dir"
+    # TODO: Remove temp_dir from folders
+    return "$status"
+}
+
+main "${@}"

src/parser/lesser_block.rs

@@ -46,7 +46,7 @@ pub fn verse_block<'b, 'g, 'r, 's>(
         ContextElement::ConsumeTrailingWhitespace(true),
         ContextElement::Context("lesser block"),
         ContextElement::ExitMatcherNode(ExitMatcherNode {
-            class: ExitClass::Beta,
+            class: ExitClass::Alpha,
             exit_matcher: &lesser_block_end_specialized,
         }),
     ];
@@ -101,7 +101,7 @@ pub fn comment_block<'b, 'g, 'r, 's>(
         ContextElement::ConsumeTrailingWhitespace(true),
         ContextElement::Context("lesser block"),
         ContextElement::ExitMatcherNode(ExitMatcherNode {
-            class: ExitClass::Beta,
+            class: ExitClass::Alpha,
             exit_matcher: &lesser_block_end_specialized,
         }),
     ];
@@ -141,7 +141,7 @@ pub fn example_block<'b, 'g, 'r, 's>(
         ContextElement::ConsumeTrailingWhitespace(true),
         ContextElement::Context("lesser block"),
         ContextElement::ExitMatcherNode(ExitMatcherNode {
-            class: ExitClass::Beta,
+            class: ExitClass::Alpha,
             exit_matcher: &lesser_block_end_specialized,
         }),
     ];
@@ -182,7 +182,7 @@ pub fn export_block<'b, 'g, 'r, 's>(
         ContextElement::ConsumeTrailingWhitespace(true),
         ContextElement::Context("lesser block"),
         ContextElement::ExitMatcherNode(ExitMatcherNode {
-            class: ExitClass::Beta,
+            class: ExitClass::Alpha,
             exit_matcher: &lesser_block_end_specialized,
         }),
     ];

src/parser/plain_link.rs

@@ -11,6 +11,7 @@ use nom::combinator::verify;
 use nom::multi::many_till;
 
 use super::org_source::OrgSource;
+use super::util::maybe_consume_object_trailing_whitespace_if_not_exiting;
 use crate::context::parser_with_context;
 use crate::context::ContextElement;
 use crate::context::ExitClass;
@@ -61,6 +62,8 @@ pub fn plain_link<'b, 'g, 'r, 's>(
     let (remaining, _separator) = tag(":")(remaining)?;
     let (remaining, path) = path_plain(context, remaining)?;
     peek(parser_with_context!(post)(context))(remaining)?;
+    let (remaining, _trailing_whitespace) =
+        maybe_consume_object_trailing_whitespace_if_not_exiting(context, remaining)?;
     let source = get_consumed(input, remaining);
     Ok((
         remaining,

src/parser/text_markup.rs

@@ -231,7 +231,8 @@ fn _text_markup_string<'b, 'g, 'r, 's, 'c>(
 ) -> Res<OrgSource<'s>, OrgSource<'s>> {
     let (remaining, _) = pre(context, input)?;
     let (remaining, open) = tag(marker_symbol)(remaining)?;
-    let (remaining, _peek_not_whitespace) = peek(not(multispace1))(remaining)?;
+    let (remaining, _peek_not_whitespace) =
+        peek(verify(anychar, |c| !c.is_whitespace() && *c != '\u{200B}'))(remaining)?;
     let text_markup_end_specialized = text_markup_end(open.into());
     let parser_context = ContextElement::ExitMatcherNode(ExitMatcherNode {
         class: ExitClass::Gamma,

src/parser/util.rs

@@ -127,15 +127,14 @@ pub fn start_of_line<'s>(input: OrgSource<'s>) -> Res<OrgSource<'s>, ()> {
 #[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
 pub fn preceded_by_whitespace<'s>(input: OrgSource<'s>) -> Res<OrgSource<'s>, ()> {
     let preceding_character = input.get_preceding_character();
-    match preceding_character {
+    if !preceding_character
-        Some('\n') | Some('\r') | Some(' ') | Some('\t') => {}
+        .map(|c| c.is_whitespace() || c == '\u{200B}') // 200B = Zero-width space
-        // If None, we are at the start of the file which is not allowed
+        .unwrap_or(false)
-        None | Some(_) => {
+    {
-            return Err(nom::Err::Error(CustomError::MyError(MyError(
+        return Err(nom::Err::Error(CustomError::MyError(MyError(
-                "Not preceded by whitespace.".into(),
+            "Not preceded by whitespace.".into(),
-            ))));
+        ))));
-        }
+    }
-    };
     Ok((input, ()))
 }