cui-llama.rn 1.7.3 → 1.7.6

package/cpp/ggml.c

@@ -61,9 +61,6 @@
 #define m512i(p) (__m512i)(p)
 #endif
 
-// precomputed f32 table for f16 (256 KB) (ggml-impl.h)
-float lm_ggml_table_f32_f16[1 << 16];
-
 #if defined(__linux__) || \
     defined(__FreeBSD__) || defined(__NetBSD__) || defined(__OpenBSD__) || \
     (defined(__APPLE__) && !TARGET_OS_TV && !TARGET_OS_WATCH)
@@ -133,7 +130,7 @@ static void lm_ggml_print_backtrace_symbols(void) {
 }
 #endif
 
-static void lm_ggml_print_backtrace(void) {
+void lm_ggml_print_backtrace(void) {
     const char * LM_GGML_NO_BACKTRACE = getenv("LM_GGML_NO_BACKTRACE");
     if (LM_GGML_NO_BACKTRACE) {
         return;
@@ -160,6 +157,10 @@ static void lm_ggml_print_backtrace(void) {
     const int parent_pid = getpid();
     const int child_pid = fork();
     if (child_pid < 0) { // error
+#if defined(__linux__)
+        close(lock[1]);
+        close(lock[0]);
+#endif
         return;
     } else if (child_pid == 0) { // child
         char attach[32];
@@ -167,6 +168,7 @@ static void lm_ggml_print_backtrace(void) {
 #if defined(__linux__)
         close(lock[1]);
         (void) !read(lock[0], lock, 1);
+        close(lock[0]);
 #endif
         // try gdb
         execlp("gdb", "gdb", "--batch",
@@ -195,7 +197,7 @@ static void lm_ggml_print_backtrace(void) {
     }
 }
 #else
-static void lm_ggml_print_backtrace(void) {
+void lm_ggml_print_backtrace(void) {
     // platform not supported
 }
 #endif
@@ -216,6 +218,8 @@ void lm_ggml_abort(const char * file, int line, const char * fmt, ...) {
     abort();
 }
 
+// lm_ggml_print_backtrace is registered with std::set_terminate by ggml.cpp
+
 //
 // logging
 //
@@ -894,12 +898,6 @@ struct lm_ggml_context {
     struct lm_ggml_object * objects_end;
 };
 
-struct lm_ggml_context_container {
-    bool used;
-
-    struct lm_ggml_context context;
-};
-
 //
 // data types
 //
@@ -967,6 +965,7 @@ static const char * LM_GGML_OP_NAME[LM_GGML_OP_COUNT] = {
     "UPSCALE",
     "PAD",
     "PAD_REFLECT_1D",
+    "ROLL",
     "ARANGE",
     "TIMESTEP_EMBEDDING",
     "ARGSORT",
@@ -997,7 +996,7 @@ static const char * LM_GGML_OP_NAME[LM_GGML_OP_COUNT] = {
     "OPT_STEP_ADAMW",
 };
 
-static_assert(LM_GGML_OP_COUNT == 82, "LM_GGML_OP_COUNT != 82");
+static_assert(LM_GGML_OP_COUNT == 83, "LM_GGML_OP_COUNT != 83");
 
 static const char * LM_GGML_OP_SYMBOL[LM_GGML_OP_COUNT] = {
     "none",
@@ -1062,6 +1061,7 @@ static const char * LM_GGML_OP_SYMBOL[LM_GGML_OP_COUNT] = {
     "upscale(x)",
     "pad(x)",
     "pad_reflect_1d(x)",
+    "roll(x)",
     "arange(start, stop, step)",
     "timestep_embedding(timesteps, dim, max_period)",
     "argsort(x)",
@@ -1092,7 +1092,7 @@ static const char * LM_GGML_OP_SYMBOL[LM_GGML_OP_COUNT] = {
     "adamw(x)",
 };
 
-static_assert(LM_GGML_OP_COUNT == 82, "LM_GGML_OP_COUNT != 82");
+static_assert(LM_GGML_OP_COUNT == 83, "LM_GGML_OP_COUNT != 83");
 
 static_assert(LM_GGML_OP_POOL_COUNT == 2, "LM_GGML_OP_POOL_COUNT != 2");
 
@@ -1432,14 +1432,6 @@ struct lm_ggml_context * lm_ggml_init(struct lm_ggml_init_params params) {
         // initialize time system (required on Windows)
         lm_ggml_time_init();
 
-        for (int i = 0; i < (1 << 16); ++i) {
-            union {
-                uint16_t u16;
-                lm_ggml_fp16_t fp16;
-            } u = {i};
-            lm_ggml_table_f32_f16[i] = LM_GGML_COMPUTE_FP16_TO_FP32(u.fp16);
-        }
-
         is_first_call = false;
     }
 
@@ -2325,6 +2317,26 @@ struct lm_ggml_tensor * lm_ggml_repeat(
     return result;
 }
 
+struct lm_ggml_tensor * lm_ggml_repeat_4d(
+        struct lm_ggml_context * ctx,
+        struct lm_ggml_tensor * a,
+        int64_t ne0, int64_t ne1, int64_t ne2, int64_t ne3) {
+    const bool can_repeat = lm_ggml_is_empty(a) || (
+        (ne0 % a->ne[0] == 0) &&
+        (ne1 % a->ne[1] == 0) &&
+        (ne2 % a->ne[2] == 0) &&
+        (ne3 % a->ne[3] == 0)
+    );
+    LM_GGML_ASSERT(can_repeat);
+
+    struct lm_ggml_tensor * result = lm_ggml_new_tensor_4d(ctx, a->type, ne0, ne1, ne2, ne3);
+
+    result->op     = LM_GGML_OP_REPEAT;
+    result->src[0] = a;
+
+    return result;
+}
+
 // lm_ggml_repeat_back
 
 struct lm_ggml_tensor * lm_ggml_repeat_back(
@@ -4333,6 +4345,34 @@ struct lm_ggml_tensor * lm_ggml_pad_reflect_1d(
     return result;
 }
 
+// lm_ggml_roll
+
+struct lm_ggml_tensor * lm_ggml_roll(
+        struct lm_ggml_context * ctx,
+        struct lm_ggml_tensor  * a,
+        int                   shift0,
+        int                   shift1,
+        int                   shift2,
+        int                   shift3) {
+    LM_GGML_ASSERT(a->nb[0] == lm_ggml_type_size(a->type));
+    LM_GGML_ASSERT(abs(shift0) < a->ne[0]);
+    LM_GGML_ASSERT(abs(shift1) < a->ne[1]);
+    LM_GGML_ASSERT(abs(shift2) < a->ne[2]);
+    LM_GGML_ASSERT(abs(shift3) < a->ne[3]);
+
+    struct lm_ggml_tensor * result = lm_ggml_dup_tensor(ctx, a);
+
+    lm_ggml_set_op_params_i32(result, 0, shift0);
+    lm_ggml_set_op_params_i32(result, 1, shift1);
+    lm_ggml_set_op_params_i32(result, 2, shift2);
+    lm_ggml_set_op_params_i32(result, 3, shift3);
+
+    result->op     = LM_GGML_OP_ROLL;
+    result->src[0] = a;
+
+    return result;
+}
+
 // lm_ggml_arange
 
 struct lm_ggml_tensor * lm_ggml_arange(

package/cpp/ggml.h

@@ -490,6 +490,7 @@ extern "C" {
         LM_GGML_OP_UPSCALE, // nearest interpolate
         LM_GGML_OP_PAD,
         LM_GGML_OP_PAD_REFLECT_1D,
+        LM_GGML_OP_ROLL,
         LM_GGML_OP_ARANGE,
         LM_GGML_OP_TIMESTEP_EMBEDDING,
         LM_GGML_OP_ARGSORT,
@@ -936,6 +937,15 @@ extern "C" {
             struct lm_ggml_tensor  * a,
             struct lm_ggml_tensor  * b);
 
+    // repeat a to the specified shape
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_repeat_4d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+                       int64_t    ne0,
+                       int64_t    ne1,
+                       int64_t    ne2,
+                       int64_t    ne3);
+
     // sums repetitions in a into shape of b
     LM_GGML_API struct lm_ggml_tensor * lm_ggml_repeat_back(
             struct lm_ggml_context * ctx,
@@ -1793,6 +1803,17 @@ extern "C" {
             int                   p0,
             int                   p1);
 
+    // Move tensor elements by an offset given for each dimension. Elements that
+    // are shifted beyond the last position are wrapped around to the beginning.
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_roll(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   shift0,
+            int                   shift1,
+            int                   shift2,
+            int                   shift3);
+
+
     // Ref: https://github.com/CompVis/stable-diffusion/blob/main/ldm/modules/diffusionmodules/util.py#L151
     // timesteps: [N,]
     // return: [N, dim]
@@ -2087,9 +2108,6 @@ extern "C" {
     LM_GGML_API struct lm_ggml_tensor * lm_ggml_graph_get_grad    (const struct lm_ggml_cgraph * cgraph, const struct lm_ggml_tensor * node);
     LM_GGML_API struct lm_ggml_tensor * lm_ggml_graph_get_grad_acc(const struct lm_ggml_cgraph * cgraph, const struct lm_ggml_tensor * node);
 
-    LM_GGML_API void                 lm_ggml_graph_export(const struct lm_ggml_cgraph * cgraph, const char * fname);
-    LM_GGML_API struct lm_ggml_cgraph * lm_ggml_graph_import(const char * fname, struct lm_ggml_context ** ctx_data, struct lm_ggml_context ** ctx_eval);
-
     // print info and performance information for the graph
     LM_GGML_API void lm_ggml_graph_print(const struct lm_ggml_cgraph * cgraph);
 
@@ -2173,6 +2191,7 @@ extern "C" {
 
     // scheduling priorities
     enum lm_ggml_sched_priority {
+        LM_GGML_SCHED_PRIO_LOW = -1,
         LM_GGML_SCHED_PRIO_NORMAL,
         LM_GGML_SCHED_PRIO_MEDIUM,
         LM_GGML_SCHED_PRIO_HIGH,

package/cpp/gguf.cpp

@@ -335,7 +335,11 @@ struct lm_gguf_context * lm_gguf_init_from_file_impl(FILE * file, struct lm_gguf
 
         for (uint32_t i = 0; i < magic.size(); i++) {
             if (magic[i] != LM_GGUF_MAGIC[i]) {
-                LM_GGML_LOG_ERROR("%s: invalid magic characters: '%c%c%c%c', expected 'GGUF'\n", __func__, magic[0], magic[1], magic[2], magic[3]);
+                char c0 = isprint(magic[0]) ? magic[0] : '?';
+                char c1 = isprint(magic[1]) ? magic[1] : '?';
+                char c2 = isprint(magic[2]) ? magic[2] : '?';
+                char c3 = isprint(magic[3]) ? magic[3] : '?';
+                LM_GGML_LOG_ERROR("%s: invalid magic characters: '%c%c%c%c', expected 'GGUF'\n", __func__, c0, c1, c2, c3);
                 lm_gguf_free(ctx);
                 return nullptr;
             }
@@ -347,11 +351,28 @@ struct lm_gguf_context * lm_gguf_init_from_file_impl(FILE * file, struct lm_gguf
     int64_t n_tensors = 0;
 
     if (ok && gr.read(ctx->version)) {
-        if (ctx->version == 1) {
+        if (ok && ctx->version == 0) {
+            LM_GGML_LOG_ERROR("%s: bad GGUF version: %" PRIu32 "\n", __func__, ctx->version);
+            ok = false;
+        }
+
+        /*
+         * bit layout is different when reading non-native endian models.
+         * assuming that the GGUF version is 3, the non-native endian model
+         * would read it as 0x30000000. we can use the AND operation against
+         * the last 4 hexadecimal digits to check if the model is the same
+         * endianness as the host system.
+        */
+        if (ok && (ctx->version & 0x0000FFFF) == 0x00000000) {
+            LM_GGML_LOG_ERROR("%s: failed to load model: this GGUF file version %" PRIu32 " is extremely large, is there a mismatch between the host and model endianness?\n", __func__, ctx->version);
+            ok = false;
+        }
+
+        if (ok && ctx->version == 1) {
             LM_GGML_LOG_ERROR("%s: GGUFv1 is no longer supported, please use a more up-to-date version\n", __func__);
             ok = false;
         }
-        if (ctx->version > LM_GGUF_VERSION) {
+        if (ok && ctx->version > LM_GGUF_VERSION) {
             LM_GGML_LOG_ERROR("%s: this GGUF file is version %" PRIu32 " but this software only supports up to version %d\n",
                 __func__, ctx->version, LM_GGUF_VERSION);
             ok = false;

package/cpp/json-partial.cpp

@@ -0,0 +1,256 @@
+#include "json-partial.h"
+
+#include "log.h"
+
+#include "nlohmann/json.hpp"
+
+#include <string>
+
+using json = nlohmann::ordered_json;
+
+enum common_json_stack_element_type {
+    COMMON_JSON_STACK_ELEMENT_OBJECT,
+    COMMON_JSON_STACK_ELEMENT_KEY,
+    COMMON_JSON_STACK_ELEMENT_ARRAY,
+};
+
+struct common_json_stack_element {
+    common_json_stack_element_type type;
+    std::string key;
+};
+
+bool common_json_parse(
+    const std::string & input,
+    const std::string & healing_marker,
+    common_json & out)
+{
+    std::string::const_iterator it = input.begin();
+    const auto end = input.end();
+    return common_json_parse(it, end, healing_marker, out);
+}
+
+bool common_json_parse(
+    std::string::const_iterator & it,
+    const std::string::const_iterator & end,
+    const std::string & healing_marker,
+    common_json & out)
+{
+    // // https://json.nlohmann.me/features/parsing/sax_interface/
+    struct json_error_locator : public nlohmann::json_sax<json> {
+        std::size_t position;
+        bool found_error;
+        std::string last_token;
+        std::string exception_message;
+        std::vector<common_json_stack_element> stack;
+
+        json_error_locator() : position(0), found_error(false) {}
+
+        bool parse_error(std::size_t position, const std::string & last_token, const json::exception & ex) override { // NOLINT
+            this->position = position - 1;
+            this->found_error = true;
+            this->last_token = last_token;
+            this->exception_message = ex.what();
+            return false;
+        }
+        void close_value() {
+            if (!stack.empty() && (stack.back().type == COMMON_JSON_STACK_ELEMENT_KEY)) {
+                stack.pop_back();
+            }
+        }
+        bool null() override { // NOLINT
+            close_value();
+            return true;
+        }
+        bool boolean(bool) override { // NOLINT
+            close_value();
+            return true;
+        }
+        bool number_integer(number_integer_t) override { // NOLINT
+            close_value();
+            return true;
+        }
+        bool number_unsigned(number_unsigned_t) override { // NOLINT
+            close_value();
+            return true;
+        }
+        bool number_float(number_float_t, const string_t &) override { // NOLINT
+            close_value();
+            return true;
+        }
+        bool string(string_t &) override { // NOLINT
+            close_value();
+            return true;
+        }
+        bool binary(binary_t &) override { // NOLINT
+            close_value();
+            return true;
+        }
+        bool start_object(std::size_t) override { // NOLINT
+            stack.push_back({COMMON_JSON_STACK_ELEMENT_OBJECT, ""});
+            return true;
+        }
+        bool end_object() override {
+            LM_GGML_ASSERT(!stack.empty() && stack.back().type == COMMON_JSON_STACK_ELEMENT_OBJECT);
+            stack.pop_back();
+            close_value();
+            return true;
+        }
+        bool key(string_t & key) override { // NOLINT
+            stack.push_back({COMMON_JSON_STACK_ELEMENT_KEY, key});
+            return true;
+        }
+        bool start_array(std::size_t) override { // NOLINT
+            stack.push_back({COMMON_JSON_STACK_ELEMENT_ARRAY, ""});
+            return true;
+        }
+        bool end_array() override {
+            LM_GGML_ASSERT(!stack.empty() && stack.back().type == COMMON_JSON_STACK_ELEMENT_ARRAY);
+            stack.pop_back();
+            close_value();
+            return true;
+        }
+    };
+    json_error_locator err_loc;
+    auto start = it;
+    json::sax_parse(it, end, &err_loc);
+
+    if (err_loc.found_error) {
+        it = start;
+        auto temptative_end = it + err_loc.position;
+        // LOG_DBG("Error at position %zu (is_end = %s): %s\n", err_loc.position, temptative_end == end ? "true" : "false", err_loc.exception_message.c_str());
+
+        auto input = std::string(it, temptative_end);
+        try {
+            out.json = json::parse(input);
+            // out.json = json::parse(it, temptative_end);
+            it = temptative_end;
+            return true;
+        } catch (const std::exception & ex) {
+            // No, needs healing.
+            LOG_DBG("Failed to parse up to error: %s: <<<%s>>>\n", ex.what(), std::string(it, temptative_end).c_str());
+        }
+        auto can_parse = [](const std::string & str) {
+            try {
+                auto _ = json::parse(str); // NOLINT
+                return true;
+            } catch (const std::exception &) {
+                return false;
+            }
+        };
+        if (!healing_marker.empty() && !err_loc.stack.empty()) {
+            std::string str(it, temptative_end);
+            auto last_non_sp_pos = str.find_last_not_of(" \n\r\t");
+            if (last_non_sp_pos == std::string::npos) {
+                throw std::runtime_error("Cannot heal a truncated JSON that stopped in an unknown location");
+            }
+            auto last_non_sp_char = str[last_non_sp_pos];
+            // Used to detect stops on a number, which may not be complete.
+            auto was_maybe_number = [&]() {
+                if (!str.empty() && std::isspace(str.back())) {
+                    return false;
+                }
+                return std::isdigit(last_non_sp_char) ||
+                    last_non_sp_char == '.' ||
+                    last_non_sp_char == 'e' ||
+                    last_non_sp_char == 'E' ||
+                    last_non_sp_char == '-';
+            };
+
+            std::string closing;
+            for (size_t i = err_loc.stack.size(); i > 0; i--) {
+                auto & el = err_loc.stack[i - 1];
+                if (el.type == COMMON_JSON_STACK_ELEMENT_OBJECT) {
+                    closing += "}";
+                } else if (el.type == COMMON_JSON_STACK_ELEMENT_ARRAY) {
+                    closing += "]";
+                } else if (el.type != COMMON_JSON_STACK_ELEMENT_KEY) {
+                    throw std::runtime_error("Unexpected stack element type");
+                }
+            }
+
+            const auto & magic_seed = out.healing_marker.marker = healing_marker;//"$llama.cpp.json$";
+
+            if (err_loc.stack.back().type == COMMON_JSON_STACK_ELEMENT_KEY) {
+                // We're inside an object value
+                if (last_non_sp_char == ':' && can_parse(str + "1" + closing)) {
+                    // Was about to create an object value
+                    str += (out.healing_marker.json_dump_marker = "\"" + magic_seed) + "\"" + closing;
+                } else if (can_parse(str + ": 1" + closing)) {
+                    str += (out.healing_marker.json_dump_marker = ":\"" + magic_seed) + "\"" + closing;
+                } else if (last_non_sp_char == '{' && can_parse(str + closing)) {
+                    // Was about to create an object
+                    str += (out.healing_marker.json_dump_marker = "\"" + magic_seed) + "\": 1" + closing;
+                } else if (can_parse(str + "\"" + closing)) {
+                    // Was inside an object value string
+                    str += (out.healing_marker.json_dump_marker = magic_seed) + "\"" + closing;
+                } else if (str[str.length() - 1] == '\\' && can_parse(str + "\\\"" + closing)) {
+                    // Was inside an object value string after an escape
+                    str += (out.healing_marker.json_dump_marker = "\\" + magic_seed) + "\"" + closing;
+                } else {
+                    // find last :
+                    auto last_pos = str.find_last_of(':');
+                    if (last_pos == std::string::npos) {
+                        throw std::runtime_error("Cannot heal a truncated JSON that stopped in an unknown location");
+                    }
+                    // Cutting back to opening : for object value
+                    str = str.substr(0, last_pos + 1) + (out.healing_marker.json_dump_marker = "\"" + magic_seed) + "\"" + closing;
+                }
+            } else if (err_loc.stack.back().type == COMMON_JSON_STACK_ELEMENT_ARRAY) {
+                if ((last_non_sp_char == ',' || last_non_sp_char == '[') && can_parse(str + "1" + closing)) {
+                    // Was about to create an array value
+                    str += (out.healing_marker.json_dump_marker = "\"" + magic_seed) + "\"" + closing;
+                } else if (can_parse(str + "\"" + closing)) {
+                    // Was inside an array value string
+                    str += (out.healing_marker.json_dump_marker = magic_seed) + "\"" + closing;
+                } else if (str[str.length() - 1] == '\\' && can_parse(str + "\\\"" + closing)) {
+                    // Was inside an array value string after an escape
+                    str += (out.healing_marker.json_dump_marker = "\\" + magic_seed) + "\"" + closing;
+                } else if (!was_maybe_number() && can_parse(str + ", 1" + closing)) {
+                    // Had just finished a value
+                    str += (out.healing_marker.json_dump_marker = ",\"" + magic_seed) + "\"" + closing;
+                } else {
+                    auto last_pos = str.find_last_of("[,");
+                    if (last_pos == std::string::npos) {
+                        throw std::runtime_error("Cannot heal a truncated JSON array stopped in an unknown location");
+                    }
+                    // Cutting back to last [ or , for array value
+                    str = str.substr(0, last_pos + 1) + (out.healing_marker.json_dump_marker = "\"" + magic_seed) + "\"" + closing;
+                }
+            } else if (err_loc.stack.back().type == COMMON_JSON_STACK_ELEMENT_OBJECT) {
+                if ((last_non_sp_char == '{' && can_parse(str + closing)) ||
+                        (last_non_sp_char == ',' && can_parse(str + "\"\": 1" + closing))) {
+                    // Was about to create an object key+value
+                    str += (out.healing_marker.json_dump_marker = "\"" + magic_seed) + "\": 1" + closing;
+                } else if (!was_maybe_number() && can_parse(str + ",\"\": 1" + closing)) {
+                    // Was about to create an object key+value
+                    str += (out.healing_marker.json_dump_marker = ",\"" + magic_seed) + "\": 1" + closing;
+                } else if (can_parse(str + "\": 1" + closing)) {
+                    // Was inside an object key string
+                    str += (out.healing_marker.json_dump_marker = magic_seed) + "\": 1" + closing;
+                } else if (str[str.length() - 1] == '\\' && can_parse(str + "\\\": 1" + closing)) {
+                    // Was inside an object key string after an escape
+                    str += (out.healing_marker.json_dump_marker = "\\" + magic_seed) + "\": 1" + closing;
+                } else {
+                    auto last_pos = str.find_last_of(':');
+                    if (last_pos == std::string::npos) {
+                        throw std::runtime_error("Cannot heal a truncated JSON object stopped in an unknown location");
+                    }
+                    // fprintf(stderr, "Cutting back to last : for object key+value\n");
+                    str = str.substr(0, last_pos + 1) + (out.healing_marker.json_dump_marker = "\"" + magic_seed) + "\"" + closing;
+                }
+            } else {
+                throw std::runtime_error("Cannot heal a truncated JSON object stopped in an unknown location");
+            }
+            // fprintf(stderr, "HEALED:\nSTRING <<<\n%s\n>>>\n\nmagic_cut: <<<\n%s\n>>>\n\n", str.c_str(), out.healing_marker.json_dump_marker.c_str());
+            out.json = json::parse(str);
+            it = temptative_end;
+            return true;
+        }
+        // TODO: handle unclosed top-level primitive if the stack was empty but we got an error (e.g. "tru", "\"", etc...)
+        // fprintf(stderr, "Closing: TODO\n");
+        return false;
+    }
+    out.json = json::parse(it, end);
+    it = end;
+    return true;
+}

package/cpp/json-partial.h

@@ -0,0 +1,38 @@
+#pragma once
+
+#include "nlohmann/json.hpp"
+
+// Healing marker (empty if the JSON was fully parsed / wasn't healed).
+struct common_healing_marker {
+    // Raw marker.
+    std::string marker;
+
+    // Cutting the `common_json.json.dump()` string at the (only) occurrence of this marker should yield the original partial JSON string (modulo spaces / if it had the same dump format).
+    std::string json_dump_marker;
+};
+
+// Represents a parsed JSON object, with its optional healing marker (a JSON dump fragment that can be used to find the position of healing in the JSON dump string)
+struct common_json {
+    nlohmann::ordered_json json;
+
+    common_healing_marker healing_marker;
+};
+
+// Parse the JSON string, healing (closing) any partial JSON if `healing_marker` is not empty.
+//
+// Healing completes partial JSON strings by adding a (possibly modified) healing marker, then whatever is needed to close the JSON.
+// This allows to parse the resulting healed JSON string, yet be able to cut it again if needed at the healing marker.
+// (this is used when parsing JSON outputs from the models, then crafting partial JSONs for the partial tool calls in OAI format).
+//
+// For instance, parsing `{` with a healing marker `foo` will produce a healed JSON `{"foo":1}`, w/ json_dump_marker = `"foo"` (which can be used to break the JSON again).
+bool common_json_parse(
+    const std::string & input,
+    const std::string & healing_marker,
+    common_json & out);
+
+// Parse the JSON string (see overload above), but advancing an iterator to the end of the input when the (potentially partial) parsing succeeds.
+bool common_json_parse(
+    std::string::const_iterator & it,
+    const std::string::const_iterator & end,
+    const std::string & healing_marker,
+    common_json & out);

package/cpp/json-schema-to-grammar.cpp

@@ -1,8 +1,9 @@
 #include "json-schema-to-grammar.h"
 #include "common.h"
 
+#include "nlohmann/json.hpp"
+
 #include <algorithm>
-#include <fstream>
 #include <map>
 #include <regex>
 #include <sstream>
@@ -40,49 +41,6 @@ static std::string build_repetition(const std::string & item_rule, int min_items
     return result;
 }
 
-/* Minimalistic replacement for std::string_view, which is only available from C++17 onwards */
-class string_view {
-    const std::string & _str;
-    const size_t _start;
-    const size_t _end;
-public:
-    string_view(const std::string & str, size_t start = 0, size_t end  = std::string::npos) : _str(str), _start(start), _end(end == std::string::npos ? str.length() : end) {}
-
-    size_t size() const {
-        return _end - _start;
-    }
-
-    size_t length() const {
-        return size();
-    }
-
-    operator std::string() const {
-        return str();
-    }
-
-    std::string str() const {
-        return _str.substr(_start, _end - _start);
-    }
-
-    string_view substr(size_t pos, size_t len = std::string::npos) const {
-        return string_view(_str, _start + pos, len == std::string::npos ? _end : _start + pos + len);
-    }
-
-    char operator[](size_t pos) const {
-        auto index = _start + pos;
-        if (index >= _end) {
-            throw std::out_of_range("string_view index out of range");
-        }
-        return _str[_start + pos];
-    }
-
-    bool operator==(const string_view & other) const {
-        std::string this_str = *this;
-        std::string other_str = other;
-        return this_str == other_str;
-    }
-};
-
 static void _build_min_max_int(int min_value, int max_value, std::stringstream & out, int decimals_left = 16, bool top_level = true) {
     auto has_min = min_value != std::numeric_limits<int>::min();
     auto has_max = max_value != std::numeric_limits<int>::max();
@@ -111,14 +69,14 @@ static void _build_min_max_int(int min_value, int max_value, std::stringstream &
         }
         out << "}";
     };
-    std::function<void(const string_view &, const string_view &)> uniform_range =
-        [&](const string_view & from, const string_view & to) {
+    std::function<void(const std::string_view &, const std::string_view &)> uniform_range =
+        [&](const std::string_view & from, const std::string_view & to) {
             size_t i = 0;
             while (i < from.length() && i < to.length() && from[i] == to[i]) {
                 i++;
             }
             if (i > 0) {
-                out << "\"" << from.substr(0, i).str() << "\"";
+                out << "\"" << from.substr(0, i) << "\"";
             }
             if (i < from.length() && i < to.length()) {
                 if (i > 0) {

package/cpp/json-schema-to-grammar.h

@@ -1,9 +1,9 @@
 #pragma once
 
-#include "ggml.h"
-// Change JSON_ASSERT from assert() to LM_GGML_ASSERT:
-#define JSON_ASSERT LM_GGML_ASSERT
-#include "json.hpp"
+#include "nlohmann/json_fwd.hpp"
+
+#include <functional>
+#include <string>
 
 std::string json_schema_to_grammar(const nlohmann::ordered_json & schema,
                                    bool force_gbnf = false);