headson 0.6.3__tar.gz → 0.6.4__tar.gz

{headson-0.6.3 → headson-0.6.4}/Cargo.lock

@@ -298,7 +298,7 @@ dependencies = [
 
 [[package]]
 name = "headson"
-version = "0.6.3"
+version = "0.6.4"
 dependencies = [
  "anyhow",
  "assert_cmd",

{headson-0.6.3 → headson-0.6.4}/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "headson"
-version = "0.6.3"
+version = "0.6.4"
 edition = "2024"
 description = "Budget‑constrained JSON preview renderer"
 readme = "README.md"

{headson-0.6.3 → headson-0.6.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: headson
-Version: 0.6.3
+Version: 0.6.4
 Classifier: Programming Language :: Python
 Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Rust
@@ -26,6 +26,9 @@ Available as:
 - CLI (see [Usage](#usage))
 - Python library (see [Python Bindings](#python-bindings))
 
+![Codecov](https://img.shields.io/codecov/c/github/kantord/headson?style=flat-square) ![Crates.io Version](https://img.shields.io/crates/v/headson?style=flat-square) ![PyPI - Version](https://img.shields.io/pypi/v/headson?style=flat-square)
+
+
 ## Install
 
 Using Cargo:
@@ -67,8 +70,8 @@ If you’re comfortable with tools like `head` and `tail`, use `headson` when yo
 
 Common flags:
 
-- `-n, --budget <BYTES>`: per‑file output budget. For multiple inputs, default total budget is `<BYTES> * number_of_inputs`.
-- `-N, --global-budget <BYTES>`: total output budget across all inputs. With `--budget`, the effective total is the smaller of the two.
+- `-c, --bytes <BYTES>`: per‑file output budget. For multiple inputs, default total budget is `<BYTES> * number_of_inputs`.
+- `-C, --global-bytes <BYTES>`: total output budget across all inputs. With `--bytes`, the effective total is the smaller of the two.
 - `-f, --format <auto|json|yaml|text>`: output format (default: `auto`).
   - Auto: stdin → JSON family; filesets → per‑file based on extension (`.json` → JSON family, `.yaml`/`.yml` → YAML, unknown → Text).
 - `-t, --template <strict|default|detailed>`: output style (default: `default`).
@@ -89,7 +92,7 @@ Notes:
   - With newlines enabled, file sections are rendered with human‑readable headers. In compact/single‑line modes, headers are omitted.
 - In `--format auto`, each file uses its own best format: JSON family for `.json`, YAML for `.yaml`/`.yml`.
   - Unknown extensions are treated as Text (raw lines) — safe for logs and `.txt` files.
-  - `--global-budget` may truncate or omit entire files to respect the total budget.
+  - `--global-bytes` may truncate or omit entire files to respect the total budget.
   - The tool finds the largest preview that fits the budget; even if extremely tight, you still get a minimal, valid preview.
   - Directories and binary files are ignored; a notice is printed to stderr for each. Stdin reads the stream as‑is.
   - Head vs Tail sampling: these options bias which part of arrays are kept before rendering. Display styles may still insert internal gap markers to honor very small budgets; strict JSON stays unannotated.
@@ -98,33 +101,33 @@ Quick one‑liners:
 
 - Peek a big JSON stream (keeps structure):
 
-      zstdcat huge.json.zst | headson -n 800 -f json -t default
+      zstdcat huge.json.zst | headson -c 800 -f json -t default
 
 - Many files with a fixed overall size:
 
-      headson -N 1200 -f json -t strict logs/*.json
+      headson -C 1200 -f json -t strict logs/*.json
 
 - Glance at a file, JavaScript‑style comments for omissions:
 
-      headson -n 400 -f json -t detailed data.json
+      headson -c 400 -f json -t detailed data.json
 
 - YAML with detailed comments:
 
-      headson -n 400 -f yaml -t detailed config.yaml
+      headson -c 400 -f yaml -t detailed config.yaml
 
 ### Text mode
 
 - Single file (auto):
 
-      headson -n 200 notes.txt
+      headson -c 200 notes.txt
 
 - Force Text ingest/output (useful when mixing with other extensions):
 
-      headson -n 200 -i text -f text notes.txt
+      headson -c 200 -i text -f text notes.txt
 
 - Many text files (fileset):
 
-      headson -n 800 -i text -f text logs/*.txt
+      headson -c 800 -i text -f text logs/*.txt
 
 - Styles on Text:
   - default: omission as a standalone `…` line.
@@ -135,6 +138,8 @@ Show help:
 
     headson --help
 
+Note: flags align with head/tail conventions (`-c/--bytes`, `-C/--global-bytes`).
+
 ## Examples: head vs headson
 
 Input:
@@ -153,7 +158,7 @@ jq -c . users.json | head -c 80
 Structured preview with headson (JSON family, default style → Pseudo):
 
 ```bash
-headson -n 120 -f json -t default users.json
+headson -c 120 -f json -t default users.json
 # {
 #   users: [
 #     { id: 1, name: "Ana", roles: [ "admin", … ] },
@@ -166,7 +171,7 @@ headson -n 120 -f json -t default users.json
 Machine‑readable preview (JSON family, strict style → strict JSON):
 
 ```bash
-headson -n 120 -f json -t strict users.json
+headson -c 120 -f json -t strict users.json
 # {"users":[{"id":1,"name":"Ana","roles":["admin"]}],"meta":{"count":2}}
 ```
 

{headson-0.6.3 → headson-0.6.4}/README.md

@@ -12,6 +12,9 @@ Available as:
 - CLI (see [Usage](#usage))
 - Python library (see [Python Bindings](#python-bindings))
 
+![Codecov](https://img.shields.io/codecov/c/github/kantord/headson?style=flat-square) ![Crates.io Version](https://img.shields.io/crates/v/headson?style=flat-square) ![PyPI - Version](https://img.shields.io/pypi/v/headson?style=flat-square)
+
+
 ## Install
 
 Using Cargo:
@@ -53,8 +56,8 @@ If you’re comfortable with tools like `head` and `tail`, use `headson` when yo
 
 Common flags:
 
-- `-n, --budget <BYTES>`: per‑file output budget. For multiple inputs, default total budget is `<BYTES> * number_of_inputs`.
-- `-N, --global-budget <BYTES>`: total output budget across all inputs. With `--budget`, the effective total is the smaller of the two.
+- `-c, --bytes <BYTES>`: per‑file output budget. For multiple inputs, default total budget is `<BYTES> * number_of_inputs`.
+- `-C, --global-bytes <BYTES>`: total output budget across all inputs. With `--bytes`, the effective total is the smaller of the two.
 - `-f, --format <auto|json|yaml|text>`: output format (default: `auto`).
   - Auto: stdin → JSON family; filesets → per‑file based on extension (`.json` → JSON family, `.yaml`/`.yml` → YAML, unknown → Text).
 - `-t, --template <strict|default|detailed>`: output style (default: `default`).
@@ -75,7 +78,7 @@ Notes:
   - With newlines enabled, file sections are rendered with human‑readable headers. In compact/single‑line modes, headers are omitted.
 - In `--format auto`, each file uses its own best format: JSON family for `.json`, YAML for `.yaml`/`.yml`.
   - Unknown extensions are treated as Text (raw lines) — safe for logs and `.txt` files.
-  - `--global-budget` may truncate or omit entire files to respect the total budget.
+  - `--global-bytes` may truncate or omit entire files to respect the total budget.
   - The tool finds the largest preview that fits the budget; even if extremely tight, you still get a minimal, valid preview.
   - Directories and binary files are ignored; a notice is printed to stderr for each. Stdin reads the stream as‑is.
   - Head vs Tail sampling: these options bias which part of arrays are kept before rendering. Display styles may still insert internal gap markers to honor very small budgets; strict JSON stays unannotated.
@@ -84,33 +87,33 @@ Quick one‑liners:
 
 - Peek a big JSON stream (keeps structure):
 
-      zstdcat huge.json.zst | headson -n 800 -f json -t default
+      zstdcat huge.json.zst | headson -c 800 -f json -t default
 
 - Many files with a fixed overall size:
 
-      headson -N 1200 -f json -t strict logs/*.json
+      headson -C 1200 -f json -t strict logs/*.json
 
 - Glance at a file, JavaScript‑style comments for omissions:
 
-      headson -n 400 -f json -t detailed data.json
+      headson -c 400 -f json -t detailed data.json
 
 - YAML with detailed comments:
 
-      headson -n 400 -f yaml -t detailed config.yaml
+      headson -c 400 -f yaml -t detailed config.yaml
 
 ### Text mode
 
 - Single file (auto):
 
-      headson -n 200 notes.txt
+      headson -c 200 notes.txt
 
 - Force Text ingest/output (useful when mixing with other extensions):
 
-      headson -n 200 -i text -f text notes.txt
+      headson -c 200 -i text -f text notes.txt
 
 - Many text files (fileset):
 
-      headson -n 800 -i text -f text logs/*.txt
+      headson -c 800 -i text -f text logs/*.txt
 
 - Styles on Text:
   - default: omission as a standalone `…` line.
@@ -121,6 +124,8 @@ Show help:
 
     headson --help
 
+Note: flags align with head/tail conventions (`-c/--bytes`, `-C/--global-bytes`).
+
 ## Examples: head vs headson
 
 Input:
@@ -139,7 +144,7 @@ jq -c . users.json | head -c 80
 Structured preview with headson (JSON family, default style → Pseudo):
 
 ```bash
-headson -n 120 -f json -t default users.json
+headson -c 120 -f json -t default users.json
 # {
 #   users: [
 #     { id: 1, name: "Ana", roles: [ "admin", … ] },
@@ -152,7 +157,7 @@ headson -n 120 -f json -t default users.json
 Machine‑readable preview (JSON family, strict style → strict JSON):
 
 ```bash
-headson -n 120 -f json -t strict users.json
+headson -c 120 -f json -t strict users.json
 # {"users":[{"id":1,"name":"Ana","roles":["admin"]}],"meta":{"count":2}}
 ```
 

{headson-0.6.3 → headson-0.6.4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "maturin"
 
 [project]
 name = "headson"
-version = "0.6.3"
+version = "0.6.4"
 description = "Budget‑constrained JSON preview renderer (Python bindings)"
 readme = "README.md"
 requires-python = ">=3.10"

{headson-0.6.3 → headson-0.6.4}/python/Cargo.lock

@@ -214,7 +214,7 @@ dependencies = [
 
 [[package]]
 name = "headson"
-version = "0.6.3"
+version = "0.6.4"
 dependencies = [
  "anyhow",
  "clap",
@@ -228,7 +228,7 @@ dependencies = [
 
 [[package]]
 name = "headson-python"
-version = "0.6.3"
+version = "0.6.4"
 dependencies = [
  "anyhow",
  "headson",

{headson-0.6.3 → headson-0.6.4}/python/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "headson-python"
-version = "0.6.3"
+version = "0.6.4"
 edition = "2021"
 publish = false
 readme = "README.md"

{headson-0.6.3/src/json_ingest → headson-0.6.4/src/ingest/formats/json}/builder.rs

@@ -5,7 +5,7 @@ use std::cell::RefCell;
 use crate::order::NodeKind;
 use crate::utils::tree_arena::{JsonTreeArena, JsonTreeNode};
 
-use super::samplers::ArraySamplerKind;
+use crate::ingest::sampling::ArraySamplerKind;
 
 #[derive(Default)]
 pub(crate) struct JsonTreeBuilder {
@@ -15,10 +15,7 @@ pub(crate) struct JsonTreeBuilder {
 }
 
 impl JsonTreeBuilder {
-    pub(crate) fn new(
-        array_cap: usize,
-        sampler: super::samplers::ArraySamplerKind,
-    ) -> Self {
+    pub(crate) fn new(array_cap: usize, sampler: ArraySamplerKind) -> Self {
         Self {
             arena: RefCell::new(JsonTreeArena::default()),
             array_cap,

{headson-0.6.3/src/json_ingest → headson-0.6.4/src/ingest/formats/json}/mod.rs

@@ -1,24 +1,27 @@
 mod builder;
 mod samplers;
-use serde::de::DeserializeSeed;
 
-use crate::PriorityConfig;
-use crate::utils::tree_arena::JsonTreeArena;
 use anyhow::Result;
 use builder::JsonTreeBuilder;
+use serde::de::DeserializeSeed;
+
+use crate::PriorityConfig;
+use crate::utils::tree_arena::JsonTreeArena as TreeArena;
+
+use crate::ingest::Ingest;
 
 #[cfg(test)]
 pub fn build_json_tree_arena(
     input: &str,
     config: &PriorityConfig,
-) -> Result<JsonTreeArena> {
+) -> Result<TreeArena> {
     build_json_tree_arena_from_bytes(input.as_bytes().to_vec(), config)
 }
 
 pub fn build_json_tree_arena_from_bytes(
     mut bytes: Vec<u8>,
     config: &PriorityConfig,
-) -> Result<JsonTreeArena> {
+) -> Result<TreeArena> {
     let mut de = simd_json::Deserializer::from_slice(&mut bytes)?;
     let builder = JsonTreeBuilder::new(
         config.array_max_items,
@@ -36,7 +39,7 @@ pub fn build_json_tree_arena_from_bytes(
 pub fn build_json_tree_arena_from_many(
     mut inputs: Vec<(String, Vec<u8>)>,
     config: &PriorityConfig,
-) -> Result<JsonTreeArena> {
+) -> Result<TreeArena> {
     let builder = JsonTreeBuilder::new(
         config.array_max_items,
         config.array_sampler.into(),
@@ -57,6 +60,38 @@ pub fn build_json_tree_arena_from_many(
     Ok(arena)
 }
 
+/// JSON adapter for the ingest boundary. Delegates to the JSON builder to
+/// produce the neutral `TreeArena`.
+pub struct JsonIngest;
+
+impl Ingest for JsonIngest {
+    fn parse_one(bytes: Vec<u8>, cfg: &PriorityConfig) -> Result<TreeArena> {
+        build_json_tree_arena_from_bytes(bytes, cfg)
+    }
+
+    fn parse_many(
+        inputs: Vec<(String, Vec<u8>)>,
+        cfg: &PriorityConfig,
+    ) -> Result<TreeArena> {
+        build_json_tree_arena_from_many(inputs, cfg)
+    }
+}
+
+/// Convenience functions for the JSON ingest path.
+pub fn parse_json_one(
+    bytes: Vec<u8>,
+    cfg: &PriorityConfig,
+) -> Result<TreeArena> {
+    JsonIngest::parse_one(bytes, cfg)
+}
+
+pub fn parse_json_many(
+    inputs: Vec<(String, Vec<u8>)>,
+    cfg: &PriorityConfig,
+) -> Result<TreeArena> {
+    JsonIngest::parse_many(inputs, cfg)
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;

{headson-0.6.3/src/json_ingest → headson-0.6.4/src/ingest/formats/json}/samplers/default.rs

@@ -1,14 +1,12 @@
 use serde::de::{IgnoredAny, SeqAccess};
 
-use super::{JsonTreeBuilder, SampledArray};
+use super::JsonTreeBuilder;
+use super::SampledArray;
 
-// Tunable sampling constants for the default strategy.
+// Default strategy phases: keep-first, greedy, then index-hash acceptance (~50%).
 const RANDOM_ACCEPT_SEED: u64 = 0x9e37_79b9_7f4a_7c15;
-// ~50% acceptance to thin remaining elements in the random phase.
-const RANDOM_ACCEPT_THRESHOLD: u32 = 0x8000_0000;
-// Keep a small, fixed number of items from the head before greedy/random phases.
+const RANDOM_ACCEPT_THRESHOLD: u32 = 0x8000_0000; // ~50%
 const KEEP_FIRST_COUNT: usize = 3;
-// Take roughly half of the remaining capacity greedily after the first items.
 const GREEDY_PORTION_DIVISOR: usize = 2;
 
 struct PhaseState {

{headson-0.6.3/src/json_ingest → headson-0.6.4/src/ingest/formats/json}/samplers/head.rs

@@ -1,6 +1,7 @@
 use serde::de::{IgnoredAny, SeqAccess};
 
-use super::{JsonTreeBuilder, SampledArray};
+use super::JsonTreeBuilder;
+use super::SampledArray;
 
 fn parse_keep<'de, A>(
     seq: &mut A,

{headson-0.6.3/src/json_ingest → headson-0.6.4/src/ingest/formats/json}/samplers/mod.rs

@@ -1,7 +1,7 @@
 use serde::de::SeqAccess;
 
-use crate::ArraySamplerStrategy;
-use crate::json_ingest::builder::JsonTreeBuilder;
+use crate::ingest::formats::json::builder::JsonTreeBuilder;
+use crate::ingest::sampling::ArraySamplerKind;
 
 #[derive(Debug)]
 pub(crate) struct SampledArray {
@@ -10,14 +10,6 @@ pub(crate) struct SampledArray {
     pub total_len: usize,
 }
 
-#[derive(Copy, Clone, Debug, Default)]
-pub(crate) enum ArraySamplerKind {
-    #[default]
-    Default,
-    Head,
-    Tail,
-}
-
 impl ArraySamplerKind {
     pub(crate) fn sample_stream<'de, A>(
         self,
@@ -38,16 +30,6 @@ impl ArraySamplerKind {
     }
 }
 
-impl From<ArraySamplerStrategy> for ArraySamplerKind {
-    fn from(strategy: ArraySamplerStrategy) -> Self {
-        match strategy {
-            ArraySamplerStrategy::Default => ArraySamplerKind::Default,
-            ArraySamplerStrategy::Head => ArraySamplerKind::Head,
-            ArraySamplerStrategy::Tail => ArraySamplerKind::Tail,
-        }
-    }
-}
-
 mod default;
 mod head;
 mod tail;

{headson-0.6.3/src/json_ingest → headson-0.6.4/src/ingest/formats/json}/samplers/tail.rs

@@ -1,6 +1,7 @@
 use serde::de::{IgnoredAny, SeqAccess};
 
-use super::{JsonTreeBuilder, SampledArray};
+use super::JsonTreeBuilder;
+use super::SampledArray;
 
 pub(crate) fn sample_stream<'de, A>(
     seq: &mut A,
@@ -89,8 +90,10 @@ mod tests {
         let mut cfg = PriorityConfig::new(usize::MAX, 5);
         cfg.array_sampler = crate::ArraySamplerStrategy::Tail;
         let arena =
-            crate::json_ingest::build_json_tree_arena_from_bytes(input, &cfg)
-                .expect("arena");
+            crate::ingest::formats::json::build_json_tree_arena_from_bytes(
+                input, &cfg,
+            )
+            .expect("arena");
         let root = &arena.nodes[arena.root_id];
         assert_eq!(root.children_len, 5, "kept 5");
         let mut orig_indices = Vec::new();

headson-0.6.4/src/ingest/formats/mod.rs

@@ -0,0 +1,9 @@
+// File-format specific ingest adapters live under this module.
+pub mod json;
+pub mod text;
+pub mod yaml;
+
+// Re-export commonly used helpers for convenience
+pub use json::{parse_json_many, parse_json_one};
+pub use text::{parse_text_many, parse_text_one};
+pub use yaml::{parse_yaml_many, parse_yaml_one};

{headson-0.6.3/src/text_ingest → headson-0.6.4/src/ingest/formats/text}/mod.rs

@@ -5,6 +5,9 @@ use crate::PriorityConfig;
 use crate::order::NodeKind;
 use crate::utils::tree_arena::{JsonTreeArena, JsonTreeNode};
 
+use crate::ingest::Ingest;
+use crate::ingest::sampling::{ArraySamplerKind, choose_indices};
+
 fn normalize_newlines(s: &str) -> Cow<'_, str> {
     // Normalize CRLF and CR to LF in a single allocation when needed.
     if s.as_bytes().contains(&b'\r') {
@@ -18,13 +21,15 @@ fn normalize_newlines(s: &str) -> Cow<'_, str> {
 struct TextArenaBuilder {
     arena: JsonTreeArena,
     array_cap: usize,
+    sampler: ArraySamplerKind,
 }
 
 impl TextArenaBuilder {
-    fn new(array_cap: usize) -> Self {
+    fn new(array_cap: usize, sampler: ArraySamplerKind) -> Self {
         Self {
             arena: JsonTreeArena::default(),
             array_cap,
+            sampler,
         }
     }
 
@@ -48,26 +53,38 @@ impl TextArenaBuilder {
 
     fn push_array_of_lines(
         &mut self,
-        lines: impl IntoIterator<Item = String>,
+        lines: &[String],
         total: usize,
     ) -> usize {
         let id = self.push_default();
-        let kept = total.min(self.array_cap);
+        let idxs = choose_indices(self.sampler, total, self.array_cap);
+        let kept = idxs.len().min(self.array_cap);
         let mut pushed = 0usize;
-        for (i, line) in lines.into_iter().enumerate() {
-            if i >= kept {
-                break;
+        for (i, &orig_index) in idxs.iter().take(kept).enumerate() {
+            if let Some(line) = lines.get(orig_index) {
+                let child = self.push_string(line.clone());
+                self.arena.children.push(child);
+                pushed = i + 1;
             }
-            let child = self.push_string(line);
-            self.arena.children.push(child);
-            pushed += 1;
         }
         let n = &mut self.arena.nodes[id];
         n.kind = NodeKind::Array;
-        // children for this array were appended after previous nodes; compute start = len(children) - pushed
         n.children_start = self.arena.children.len().saturating_sub(pushed);
         n.children_len = pushed;
         n.array_len = Some(total);
+        // Store arr_indices when not contiguous head prefix
+        let contiguous =
+            idxs.iter().take(kept).enumerate().all(|(i, &idx)| i == idx);
+        if pushed == 0 || contiguous {
+            n.arr_indices_start = 0;
+            n.arr_indices_len = 0;
+        } else {
+            let start = self.arena.arr_indices.len();
+            self.arena.arr_indices.extend(idxs.into_iter().take(kept));
+            let len = self.arena.arr_indices.len().saturating_sub(start);
+            n.arr_indices_start = start;
+            n.arr_indices_len = len.min(pushed);
+        }
         id
     }
 
@@ -110,8 +127,11 @@ pub fn build_text_tree_arena_from_bytes(
         .map(std::string::ToString::to_string)
         .collect();
     let total = lines_vec.len();
-    let mut b = TextArenaBuilder::new(config.array_max_items);
-    let root_id = b.push_array_of_lines(lines_vec, total);
+    let mut b = TextArenaBuilder::new(
+        config.array_max_items,
+        config.array_sampler.into(),
+    );
+    let root_id = b.push_array_of_lines(&lines_vec, total);
     let mut a = b.finish();
     a.root_id = root_id;
     Ok(a)
@@ -125,7 +145,10 @@ pub fn build_text_tree_arena_from_many(
     mut inputs: Vec<(String, Vec<u8>)>,
     config: &PriorityConfig,
 ) -> Result<JsonTreeArena> {
-    let mut b = TextArenaBuilder::new(config.array_max_items);
+    let mut b = TextArenaBuilder::new(
+        config.array_max_items,
+        config.array_sampler.into(),
+    );
     let mut keys: Vec<String> = Vec::with_capacity(inputs.len());
     let mut children_ids: Vec<usize> = Vec::with_capacity(inputs.len());
     for (key, bytes) in inputs.drain(..) {
@@ -136,7 +159,7 @@ pub fn build_text_tree_arena_from_many(
             .map(std::string::ToString::to_string)
             .collect();
         let total = lines_vec.len();
-        let child_id = b.push_array_of_lines(lines_vec, total);
+        let child_id = b.push_array_of_lines(&lines_vec, total);
         keys.push(key);
         children_ids.push(child_id);
     }
@@ -147,6 +170,39 @@ pub fn build_text_tree_arena_from_many(
     Ok(a)
 }
 
+pub struct TextIngest;
+
+impl Ingest for TextIngest {
+    fn parse_one(
+        bytes: Vec<u8>,
+        cfg: &PriorityConfig,
+    ) -> Result<JsonTreeArena> {
+        build_text_tree_arena_from_bytes(bytes, cfg)
+    }
+
+    fn parse_many(
+        inputs: Vec<(String, Vec<u8>)>,
+        cfg: &PriorityConfig,
+    ) -> Result<JsonTreeArena> {
+        build_text_tree_arena_from_many(inputs, cfg)
+    }
+}
+
+/// Convenience functions for the Text ingest path.
+pub fn parse_text_one(
+    bytes: Vec<u8>,
+    cfg: &PriorityConfig,
+) -> Result<JsonTreeArena> {
+    TextIngest::parse_one(bytes, cfg)
+}
+
+pub fn parse_text_many(
+    inputs: Vec<(String, Vec<u8>)>,
+    cfg: &PriorityConfig,
+) -> Result<JsonTreeArena> {
+    TextIngest::parse_many(inputs, cfg)
+}
+
 #[cfg(test)]
 mod tests {
     use crate::{
@@ -189,4 +245,30 @@ mod tests {
         let out = headson_text(input.into_bytes(), &cfg, &prio, 20).unwrap();
         assert!(out.contains("…\n"));
     }
+
+    #[test]
+    fn tail_sampler_keeps_last_n_indices_text() {
+        // Build 10 lines; with array_max_items=5 and tail sampler we should keep last 5
+        let lines = (0..10)
+            .map(|i| i.to_string())
+            .collect::<Vec<_>>()
+            .join("\n");
+        let mut cfg = PriorityConfig::new(usize::MAX, 5);
+        cfg.array_sampler = crate::ArraySamplerStrategy::Tail;
+        let arena =
+            super::build_text_tree_arena_from_bytes(lines.into_bytes(), &cfg)
+                .expect("arena");
+        let root = &arena.nodes[arena.root_id];
+        assert_eq!(root.children_len, 5, "kept 5");
+        let mut orig_indices = Vec::new();
+        for i in 0..root.children_len {
+            let oi = if root.arr_indices_len > 0 {
+                arena.arr_indices[root.arr_indices_start + i]
+            } else {
+                i
+            };
+            orig_indices.push(oi);
+        }
+        assert_eq!(orig_indices, vec![5, 6, 7, 8, 9]);
+    }
 }