tranfi 0.1.2 → 0.2.0

package/csrc/op_window.c

@@ -11,28 +11,142 @@
 #include <string.h>
 #include <stdio.h>
 #include <math.h>
+#include <ctype.h>
 
 typedef enum {
     WIN_AVG, WIN_SUM, WIN_MIN, WIN_MAX, WIN_COUNT,
 } win_func;
 
+typedef enum {
+    WIN_MISSING_ERROR,
+    WIN_MISSING_NULL,
+    WIN_MISSING_IGNORE,
+} window_missing_policy;
+
+typedef enum {
+    WIN_TYPE_FAIL,
+    WIN_TYPE_NULL,
+} window_type_policy;
+
 typedef struct {
     char    *column;
     char    *result;
+    char    *op_label;
     win_func func;
     size_t   size;
     double  *ring;     /* circular buffer */
     size_t   head;     /* next write position */
     size_t   count;    /* items in buffer */
+    window_missing_policy missing;
+    window_type_policy on_type_error;
 } window_state;
 
-static win_func parse_win_func(const char *s) {
-    if (strcmp(s, "avg") == 0) return WIN_AVG;
-    if (strcmp(s, "sum") == 0) return WIN_SUM;
-    if (strcmp(s, "min") == 0) return WIN_MIN;
-    if (strcmp(s, "max") == 0) return WIN_MAX;
-    if (strcmp(s, "count") == 0) return WIN_COUNT;
-    return WIN_AVG;
+static int parse_win_func(const char *s, win_func *out) {
+    if (strcmp(s, "avg") == 0) { *out = WIN_AVG; return TF_OK; }
+    if (strcmp(s, "sum") == 0) { *out = WIN_SUM; return TF_OK; }
+    if (strcmp(s, "min") == 0) { *out = WIN_MIN; return TF_OK; }
+    if (strcmp(s, "max") == 0) { *out = WIN_MAX; return TF_OK; }
+    if (strcmp(s, "count") == 0) { *out = WIN_COUNT; return TF_OK; }
+    return TF_ERROR;
+}
+
+static int window_is_numeric_type(tf_type type) {
+    return type == TF_TYPE_INT64 || type == TF_TYPE_FLOAT64;
+}
+
+static double window_get_numeric(const tf_batch *b, size_t r, int ci) {
+    if (b->col_types[ci] == TF_TYPE_INT64) return (double)tf_batch_get_int64(b, r, ci);
+    return tf_batch_get_float64(b, r, ci);
+}
+
+static const char *window_label(const window_state *st) {
+    return (st && st->op_label) ? st->op_label : "window";
+}
+
+static char *window_default_result_name(const char *column, const char *func, size_t size) {
+    char size_buf[32];
+    int n = snprintf(size_buf, sizeof(size_buf), "%zu", size);
+    if (n < 0 || (size_t)n >= sizeof(size_buf)) return NULL;
+
+    char *func_suffix = tf_string_append_suffix_checked("_", func);
+    if (!func_suffix) return NULL;
+    char *full_suffix = tf_string_append_suffix_checked(func_suffix, size_buf);
+    free(func_suffix);
+    if (!full_suffix) return NULL;
+
+    char *result = tf_string_append_suffix_checked(column, full_suffix);
+    free(full_suffix);
+    return result;
+}
+
+static void window_set_col_error(const window_state *st, const char *suffix) {
+    char msg[512];
+    snprintf(msg, sizeof(msg), "%s: column '%s' %s", window_label(st), st && st->column ? st->column : "", suffix);
+    tf_set_last_error(msg);
+}
+
+static int window_parse_missing_policy(const cJSON *args, const char *label, window_missing_policy *out) {
+    *out = WIN_MISSING_ERROR;
+    const cJSON *j = cJSON_GetObjectItemCaseSensitive(args, "missing");
+    if (!j) return TF_OK;
+    if (!cJSON_IsString(j)) {
+        char msg[128];
+        snprintf(msg, sizeof(msg), "%s: missing must be error, null, or ignore", label);
+        tf_set_last_error(msg);
+        return TF_ERROR;
+    }
+    if (strcmp(j->valuestring, "error") == 0) *out = WIN_MISSING_ERROR;
+    else if (strcmp(j->valuestring, "null") == 0) *out = WIN_MISSING_NULL;
+    else if (strcmp(j->valuestring, "ignore") == 0) *out = WIN_MISSING_IGNORE;
+    else {
+        char msg[128];
+        snprintf(msg, sizeof(msg), "%s: missing must be error, null, or ignore", label);
+        tf_set_last_error(msg);
+        return TF_ERROR;
+    }
+    return TF_OK;
+}
+
+static int window_parse_type_policy(const cJSON *args, const char *label, window_type_policy *out) {
+    *out = WIN_TYPE_FAIL;
+    const cJSON *j = cJSON_GetObjectItemCaseSensitive(args, "on_type_error");
+    if (!j) return TF_OK;
+    if (!cJSON_IsString(j)) {
+        char msg[128];
+        snprintf(msg, sizeof(msg), "%s: on_type_error must be fail or null", label);
+        tf_set_last_error(msg);
+        return TF_ERROR;
+    }
+    if (strcmp(j->valuestring, "fail") == 0) *out = WIN_TYPE_FAIL;
+    else if (strcmp(j->valuestring, "null") == 0) *out = WIN_TYPE_NULL;
+    else {
+        char msg[128];
+        snprintf(msg, sizeof(msg), "%s: on_type_error must be fail or null", label);
+        tf_set_last_error(msg);
+        return TF_ERROR;
+    }
+    return TF_OK;
+}
+
+static int window_passthrough(tf_batch *in, tf_batch **out) {
+    tf_batch *ob = tf_batch_create(in->n_cols, in->n_rows);
+    if (!ob) return TF_ERROR;
+    if (tf_batch_clone_schema(ob, in) != TF_OK) {
+        tf_batch_free(ob);
+        return TF_ERROR;
+    }
+    for (size_t r = 0; r < in->n_rows; r++) {
+        if (tf_batch_copy_row(ob, r, in, r) != TF_OK) {
+            tf_batch_free(ob);
+            return TF_ERROR;
+        }
+        if (tf_batch_expose_row(ob, r) != TF_OK) {
+            tf_batch_free(ob);
+            return TF_ERROR;
+        }
+    }
+    *out = ob;
+    return TF_OK;
 }
 
 static int window_process(tf_step *self, tf_batch *in, tf_batch **out,
@@ -41,31 +155,47 @@ static int window_process(tf_step *self, tf_batch *in, tf_batch **out,
     window_state *st = self->state;
     *out = NULL;
 
+    int ci = tf_batch_col_index(in, st->column);
+    int force_null = 0;
+    if (ci < 0) {
+        if (st->missing == WIN_MISSING_ERROR) {
+            window_set_col_error(st, "not found");
+            return TF_ERROR;
+        }
+        if (st->missing == WIN_MISSING_IGNORE) return window_passthrough(in, out);
+        force_null = 1;
+    } else if (st->func != WIN_COUNT && !window_is_numeric_type(in->col_types[ci])) {
+        if (st->on_type_error == WIN_TYPE_FAIL) {
+            window_set_col_error(st, "must be numeric");
+            return TF_ERROR;
+        }
+        force_null = 1;
+    }
+
+    const char *extra_names[1] = {st->result};
+    tf_type extra_types[1] = {TF_TYPE_FLOAT64};
     tf_batch *ob = tf_batch_create(in->n_cols + 1, in->n_rows);
     if (!ob) return TF_ERROR;
-    for (size_t c = 0; c < in->n_cols; c++)
-        tf_batch_set_schema(ob, c, in->col_names[c], in->col_types[c]);
-    tf_batch_set_schema(ob, in->n_cols, st->result, TF_TYPE_FLOAT64);
-
-    int ci = tf_batch_col_index(in, st->column);
+    if (tf_batch_clone_with_extra_cols(ob, in, extra_names, extra_types, 1) != TF_OK) {
+        tf_batch_free(ob);
+        return TF_ERROR;
+    }
 
     for (size_t r = 0; r < in->n_rows; r++) {
-        tf_batch_copy_row(ob, r, in, r);
+        if (tf_batch_copy_row(ob, r, in, r) != TF_OK) {
+            tf_batch_free(ob);
+            return TF_ERROR;
+        }
 
-        if (ci < 0 || tf_batch_is_null(in, r, ci)) {
-            tf_batch_set_null(ob, r, in->n_cols);
-            ob->n_rows = r + 1;
+        if (force_null || tf_batch_is_null(in, r, ci)) {
+            if (tf_batch_set_null(ob, r, in->n_cols) != TF_OK) { tf_batch_free(ob); return TF_ERROR; }
+            if (tf_batch_expose_row(ob, r) != TF_OK) { tf_batch_free(ob); return TF_ERROR; }
             continue;
         }
 
-        double val = 0;
-        if (in->col_types[ci] == TF_TYPE_INT64) val = (double)tf_batch_get_int64(in, r, ci);
-        else if (in->col_types[ci] == TF_TYPE_FLOAT64) val = tf_batch_get_float64(in, r, ci);
-
-        /* Add to ring buffer */
-        st->ring[st->head] = val;
-        st->head = (st->head + 1) % st->size;
-        if (st->count < st->size) st->count++;
+        double val = st->func == WIN_COUNT ? 0.0 : window_get_numeric(in, r, ci);
+        size_t next_head = (st->head + 1) % st->size;
+        size_t next_count = st->count < st->size ? st->count + 1 : st->count;
 
         /* Compute window aggregate */
         double result = 0;
@@ -73,29 +203,43 @@ static int window_process(tf_step *self, tf_batch *in, tf_batch **out,
             case WIN_SUM:
             case WIN_AVG: {
                 double sum = 0;
-                for (size_t i = 0; i < st->count; i++) sum += st->ring[i];
-                result = (st->func == WIN_AVG) ? sum / st->count : sum;
+                for (size_t i = 0; i < next_count; i++)
+                    sum += i == st->head ? val : st->ring[i];
+                result = (st->func == WIN_AVG) ? sum / next_count : sum;
                 break;
             }
             case WIN_MIN: {
-                result = st->ring[0];
-                for (size_t i = 1; i < st->count; i++)
-                    if (st->ring[i] < result) result = st->ring[i];
+                result = (st->head == 0) ? val : st->ring[0];
+                for (size_t i = 1; i < next_count; i++) {
+                    double item = i == st->head ? val : st->ring[i];
+                    if (item < result) result = item;
+                }
                 break;
             }
             case WIN_MAX: {
-                result = st->ring[0];
-                for (size_t i = 1; i < st->count; i++)
-                    if (st->ring[i] > result) result = st->ring[i];
+                result = (st->head == 0) ? val : st->ring[0];
+                for (size_t i = 1; i < next_count; i++) {
+                    double item = i == st->head ? val : st->ring[i];
+                    if (item > result) result = item;
+                }
                 break;
             }
             case WIN_COUNT:
-                result = (double)st->count;
+                result = (double)next_count;
                 break;
         }
 
-        tf_batch_set_float64(ob, r, in->n_cols, result);
-        ob->n_rows = r + 1;
+        if (tf_batch_set_float64(ob, r, in->n_cols, result) != TF_OK) {
+            tf_batch_free(ob);
+            return TF_ERROR;
+        }
+        if (tf_batch_expose_row(ob, r) != TF_OK) {
+            tf_batch_free(ob);
+            return TF_ERROR;
+        }
+        st->ring[st->head] = val;
+        st->head = next_head;
+        st->count = next_count;
     }
 
     *out = ob;
@@ -108,40 +252,337 @@ static int window_flush(tf_step *self, tf_batch **out, tf_side_channels *side) {
 
 static void window_destroy(tf_step *self) {
     window_state *st = self->state;
-    if (st) { free(st->column); free(st->result); free(st->ring); free(st); }
+    if (st) { free(st->column); free(st->result); free(st->op_label); free(st->ring); free(st); }
+    free(self);
+}
+
+
+static tf_step *tf_window_create_alias(const cJSON *args, const char *func, const char *label) {
+    if (!args || !func || !label) return NULL;
+    cJSON *copy = cJSON_Duplicate(args, 1);
+    if (!copy) return NULL;
+    cJSON_DeleteItemFromObjectCaseSensitive(copy, "func");
+    cJSON_DeleteItemFromObjectCaseSensitive(copy, "__op_label");
+    if (tf_json_add_string(copy, "__op_label", label) != TF_OK ||
+        tf_json_add_string(copy, "func", func) != TF_OK) {
+        cJSON_Delete(copy);
+        return NULL;
+    }
+    tf_step *step = tf_window_create(copy);
+    cJSON_Delete(copy);
+    return step;
+}
+
+tf_step *tf_rolling_sum_create(const cJSON *args) {
+    return tf_window_create_alias(args, "sum", "rolling-sum");
+}
+
+tf_step *tf_rolling_mean_create(const cJSON *args) {
+    return tf_window_create_alias(args, "avg", "rolling-mean");
+}
+
+tf_step *tf_rolling_min_create(const cJSON *args) {
+    return tf_window_create_alias(args, "min", "rolling-min");
+}
+
+tf_step *tf_rolling_max_create(const cJSON *args) {
+    return tf_window_create_alias(args, "max", "rolling-max");
+}
+
+typedef enum {
+    BOOL_ROLL_ANY,
+    BOOL_ROLL_ALL,
+} bool_roll_func;
+
+typedef enum {
+    BOOL_NULLS_IGNORE,
+    BOOL_NULLS_FALSE,
+    BOOL_NULLS_TRUE,
+    BOOL_NULLS_PROPAGATE,
+} bool_null_policy;
+
+typedef struct {
+    char            *column;
+    char            *result;
+    bool_roll_func   func;
+    bool_null_policy nulls;
+    size_t           size;
+    uint8_t         *values;
+    uint8_t         *is_null;
+    size_t           head;
+    size_t           count;
+} bool_window_state;
+
+static int streq_ci(const char *a, const char *b) {
+    if (!a || !b) return 0;
+    while (*a && *b) {
+        unsigned char ca = (unsigned char)*a;
+        unsigned char cb = (unsigned char)*b;
+        if (tolower(ca) != tolower(cb)) return 0;
+        a++;
+        b++;
+    }
+    return *a == '\0' && *b == '\0';
+}
+
+static int parse_bool_null_policy(const cJSON *args, bool_null_policy *out) {
+    *out = BOOL_NULLS_IGNORE;
+    cJSON *nulls_j = cJSON_GetObjectItemCaseSensitive(args, "nulls");
+    if (!nulls_j) return TF_OK;
+    if (!cJSON_IsString(nulls_j)) {
+        tf_set_last_error("rolling-any/all: nulls must be ignore, false, true, or propagate");
+        return TF_ERROR;
+    }
+    const char *s = nulls_j->valuestring;
+    if (strcmp(s, "ignore") == 0) *out = BOOL_NULLS_IGNORE;
+    else if (strcmp(s, "false") == 0) *out = BOOL_NULLS_FALSE;
+    else if (strcmp(s, "true") == 0) *out = BOOL_NULLS_TRUE;
+    else if (strcmp(s, "propagate") == 0) *out = BOOL_NULLS_PROPAGATE;
+    else {
+        tf_set_last_error("rolling-any/all: nulls must be ignore, false, true, or propagate");
+        return TF_ERROR;
+    }
+    return TF_OK;
+}
+
+static int cell_to_bool(const tf_batch *b, size_t row, int col, uint8_t *out) {
+    if (col < 0 || tf_batch_is_null(b, row, (size_t)col)) return 0;
+    switch (b->col_types[col]) {
+        case TF_TYPE_BOOL:
+            *out = tf_batch_get_bool(b, row, (size_t)col) ? 1 : 0;
+            return 1;
+        case TF_TYPE_INT64:
+            *out = tf_batch_get_int64(b, row, (size_t)col) != 0 ? 1 : 0;
+            return 1;
+        case TF_TYPE_FLOAT64:
+            *out = tf_batch_get_float64(b, row, (size_t)col) != 0.0 ? 1 : 0;
+            return 1;
+        case TF_TYPE_STRING: {
+            const char *s = tf_batch_get_string(b, row, (size_t)col);
+            if (!s) return 0;
+            if (streq_ci(s, "true") || streq_ci(s, "t") || streq_ci(s, "yes") ||
+                streq_ci(s, "y") || strcmp(s, "1") == 0) {
+                *out = 1;
+                return 1;
+            }
+            if (streq_ci(s, "false") || streq_ci(s, "f") || streq_ci(s, "no") ||
+                streq_ci(s, "n") || strcmp(s, "0") == 0 || strcmp(s, "") == 0) {
+                *out = 0;
+                return 1;
+            }
+            return 0;
+        }
+        default:
+            return 0;
+    }
+}
+
+static int bool_window_process(tf_step *self, tf_batch *in, tf_batch **out,
+                               tf_side_channels *side) {
+    (void)side;
+    bool_window_state *st = self->state;
+    *out = NULL;
+
+    const char *extra_names[1] = {st->result};
+    tf_type extra_types[1] = {TF_TYPE_BOOL};
+    tf_batch *ob = tf_batch_create(in->n_cols + 1, in->n_rows);
+    if (!ob) return TF_ERROR;
+    if (tf_batch_clone_with_extra_cols(ob, in, extra_names, extra_types, 1) != TF_OK) {
+        tf_batch_free(ob);
+        return TF_ERROR;
+    }
+
+    int ci = tf_batch_col_index(in, st->column);
+
+    for (size_t r = 0; r < in->n_rows; r++) {
+        if (tf_batch_copy_row(ob, r, in, r) != TF_OK) {
+            tf_batch_free(ob);
+            return TF_ERROR;
+        }
+
+        uint8_t value = 0;
+        uint8_t is_null = cell_to_bool(in, r, ci, &value) ? 0 : 1;
+        size_t next_head = (st->head + 1) % st->size;
+        size_t next_count = st->count < st->size ? st->count + 1 : st->count;
+
+        size_t non_null = 0;
+        int seen_null = 0;
+        int any_true = 0;
+        int any_false = 0;
+        for (size_t i = 0; i < next_count; i++) {
+            uint8_t item_null = i == st->head ? is_null : st->is_null[i];
+            uint8_t item_value = i == st->head ? value : st->values[i];
+            if (item_null) {
+                seen_null = 1;
+                continue;
+            }
+            non_null++;
+            if (item_value) any_true = 1;
+            else any_false = 1;
+        }
+
+        int out_null = 0;
+        int out_value = 0;
+        switch (st->nulls) {
+            case BOOL_NULLS_IGNORE:
+                if (non_null == 0) out_null = 1;
+                else out_value = (st->func == BOOL_ROLL_ANY) ? any_true : !any_false;
+                break;
+            case BOOL_NULLS_FALSE:
+                out_value = (st->func == BOOL_ROLL_ANY) ? any_true : (!seen_null && !any_false);
+                break;
+            case BOOL_NULLS_TRUE:
+                out_value = (st->func == BOOL_ROLL_ANY) ? (seen_null || any_true) : !any_false;
+                break;
+            case BOOL_NULLS_PROPAGATE:
+                if (seen_null) out_null = 1;
+                else out_value = (st->func == BOOL_ROLL_ANY) ? any_true : !any_false;
+                break;
+        }
+
+        int write_rc = out_null ? tf_batch_set_null(ob, r, in->n_cols)
+                                : tf_batch_set_bool(ob, r, in->n_cols, out_value != 0);
+        if (write_rc != TF_OK) {
+            tf_batch_free(ob);
+            return TF_ERROR;
+        }
+        if (tf_batch_expose_row(ob, r) != TF_OK) {
+            tf_batch_free(ob);
+            return TF_ERROR;
+        }
+        st->values[st->head] = value;
+        st->is_null[st->head] = is_null;
+        st->head = next_head;
+        st->count = next_count;
+    }
+
+    *out = ob;
+    return TF_OK;
+}
+
+static int bool_window_flush(tf_step *self, tf_batch **out, tf_side_channels *side) {
+    (void)self; (void)side; *out = NULL; return TF_OK;
+}
+
+static void bool_window_destroy(tf_step *self) {
+    bool_window_state *st = self->state;
+    if (st) {
+        free(st->column);
+        free(st->result);
+        free(st->values);
+        free(st->is_null);
+        free(st);
+    }
     free(self);
 }
 
+static tf_step *tf_rolling_bool_create(const cJSON *args, bool_roll_func func) {
+    if (!args) return NULL;
+    cJSON *col_j = cJSON_GetObjectItemCaseSensitive(args, "column");
+    size_t win_size = 0;
+    int has_size = tf_json_get_size_arg(args, "size",
+                                        1, TF_MAX_WINDOW_SIZE,
+                                        &win_size, "rolling-any/all");
+    if (has_size < 0) return NULL;
+    if (!cJSON_IsString(col_j) || has_size == 0) {
+        tf_set_last_error("rolling-any/all: requires column and positive size");
+        return NULL;
+    }
+
+    bool_null_policy nulls;
+    if (parse_bool_null_policy(args, &nulls) != TF_OK) return NULL;
+
+    bool_window_state *st = tf_callocarray_checked(1, sizeof(bool_window_state));
+    if (!st) return NULL;
+    st->column = tf_strdup_checked(col_j->valuestring);
+    st->func = func;
+    st->nulls = nulls;
+    st->size = win_size;
+    st->values = tf_callocarray_checked(win_size, sizeof(uint8_t));
+    st->is_null = tf_callocarray_checked(win_size, sizeof(uint8_t));
+
+    cJSON *res_j = cJSON_GetObjectItemCaseSensitive(args, "result");
+    if (cJSON_IsString(res_j)) {
+        st->result = tf_strdup_checked(res_j->valuestring);
+    } else {
+        st->result = window_default_result_name(st->column,
+                                                func == BOOL_ROLL_ANY ? "any" : "all",
+                                                win_size);
+    }
+
+    if (!st->column || !st->result || !st->values || !st->is_null) {
+        free(st->column); free(st->result); free(st->values); free(st->is_null); free(st);
+        return NULL;
+    }
+
+    tf_step *step = tf_callocarray_checked(1, sizeof(tf_step));
+    if (!step) { free(st->column); free(st->result); free(st->values); free(st->is_null); free(st); return NULL; }
+    step->process = bool_window_process;
+    step->flush = bool_window_flush;
+    step->destroy = bool_window_destroy;
+    step->state = st;
+    return step;
+}
+
+tf_step *tf_rolling_any_create(const cJSON *args) {
+    return tf_rolling_bool_create(args, BOOL_ROLL_ANY);
+}
+
+tf_step *tf_rolling_all_create(const cJSON *args) {
+    return tf_rolling_bool_create(args, BOOL_ROLL_ALL);
+}
+
 tf_step *tf_window_create(const cJSON *args) {
     if (!args) return NULL;
     cJSON *col_j = cJSON_GetObjectItemCaseSensitive(args, "column");
-    cJSON *size_j = cJSON_GetObjectItemCaseSensitive(args, "size");
     cJSON *func_j = cJSON_GetObjectItemCaseSensitive(args, "func");
-    if (!cJSON_IsString(col_j) || !cJSON_IsNumber(size_j) || !cJSON_IsString(func_j))
+    cJSON *label_j = cJSON_GetObjectItemCaseSensitive(args, "__op_label");
+    const char *label = cJSON_IsString(label_j) ? label_j->valuestring : "window";
+    size_t win_size = 0;
+    int has_size = tf_json_get_size_arg(args, "size",
+                                        1, TF_MAX_WINDOW_SIZE,
+                                        &win_size, label);
+    if (has_size < 0) return NULL;
+    if (!cJSON_IsString(col_j) || !col_j->valuestring[0] || has_size == 0 || !cJSON_IsString(func_j)) {
+        char msg[128];
+        snprintf(msg, sizeof(msg), "%s: column, size, and func are required", label);
+        tf_set_last_error(msg);
         return NULL;
+    }
 
-    size_t win_size = (size_t)size_j->valueint;
-    if (win_size == 0) win_size = 1;
+    win_func func;
+    if (parse_win_func(func_j->valuestring, &func) != TF_OK) {
+        char msg[128];
+        snprintf(msg, sizeof(msg), "%s: func must be avg, sum, min, max, or count", label);
+        tf_set_last_error(msg);
+        return NULL;
+    }
 
-    window_state *st = calloc(1, sizeof(window_state));
+    window_state *st = tf_callocarray_checked(1, sizeof(window_state));
     if (!st) return NULL;
-    st->column = strdup(col_j->valuestring);
-    st->func = parse_win_func(func_j->valuestring);
+    st->column = tf_strdup_checked(col_j->valuestring);
+    st->op_label = tf_strdup_checked(label);
+    st->func = func;
     st->size = win_size;
-    st->ring = calloc(win_size, sizeof(double));
+    st->ring = tf_callocarray_checked(win_size, sizeof(double));
+    if (!st->column || !st->op_label || !st->ring) {
+        free(st->column); free(st->op_label); free(st->ring); free(st); return NULL;
+    }
+    if (window_parse_missing_policy(args, label, &st->missing) != TF_OK ||
+        window_parse_type_policy(args, label, &st->on_type_error) != TF_OK) {
+        free(st->column); free(st->op_label); free(st->ring); free(st); return NULL;
+    }
 
     cJSON *res_j = cJSON_GetObjectItemCaseSensitive(args, "result");
     if (cJSON_IsString(res_j)) {
-        st->result = strdup(res_j->valuestring);
+        st->result = tf_strdup_checked(res_j->valuestring);
     } else {
-        char buf[256];
-        snprintf(buf, sizeof(buf), "%s_%s%zu", st->column,
-                 func_j->valuestring, win_size);
-        st->result = strdup(buf);
+        st->result = window_default_result_name(st->column, func_j->valuestring, win_size);
     }
+    if (!st->result) { free(st->column); free(st->op_label); free(st->ring); free(st); return NULL; }
 
-    tf_step *step = malloc(sizeof(tf_step));
-    if (!step) { free(st->column); free(st->result); free(st->ring); free(st); return NULL; }
+    tf_step *step = tf_callocarray_checked(1, sizeof(tf_step));
+    if (!step) { free(st->column); free(st->result); free(st->op_label); free(st->ring); free(st); return NULL; }
     step->process = window_process;
     step->flush = window_flush;
     step->destroy = window_destroy;