455 lines
14 KiB
C++
455 lines
14 KiB
C++
#include <fesa/validation/reference_csv.hpp>
|
|
|
|
#include <algorithm>
|
|
#include <array>
|
|
#include <charconv>
|
|
#include <cmath>
|
|
#include <cstddef>
|
|
#include <cstdint>
|
|
#include <fstream>
|
|
#include <iterator>
|
|
#include <map>
|
|
#include <optional>
|
|
#include <set>
|
|
#include <sstream>
|
|
#include <string>
|
|
#include <string_view>
|
|
#include <tuple>
|
|
#include <utility>
|
|
#include <vector>
|
|
|
|
namespace fesa {
|
|
namespace {
|
|
|
|
constexpr std::string_view instance_column{"Part Instance Name"};
|
|
constexpr std::string_view node_column{"Node Label"};
|
|
constexpr std::string_view element_column{"Element Label"};
|
|
|
|
using PositionKey =
|
|
std::tuple<std::string, std::int64_t, std::optional<std::int64_t>>;
|
|
using ColumnIndices =
|
|
std::map<std::string, std::size_t, std::less<>>;
|
|
|
|
std::size_t column_index(
|
|
const ColumnIndices& column_indices,
|
|
const std::string_view name) {
|
|
return column_indices.find(name)->second;
|
|
}
|
|
|
|
std::string_view trim(const std::string_view value) {
|
|
constexpr std::string_view whitespace{" \t\f\v\r\n"};
|
|
const std::size_t first = value.find_first_not_of(whitespace);
|
|
if (first == std::string_view::npos) {
|
|
return {};
|
|
}
|
|
const std::size_t last = value.find_last_not_of(whitespace);
|
|
return value.substr(first, last - first + 1U);
|
|
}
|
|
|
|
std::vector<std::string> split_fields(const std::string_view line) {
|
|
std::vector<std::string> fields;
|
|
std::size_t first = 0U;
|
|
while (true) {
|
|
const std::size_t comma = line.find(',', first);
|
|
const std::string_view field =
|
|
comma == std::string_view::npos
|
|
? line.substr(first)
|
|
: line.substr(first, comma - first);
|
|
fields.emplace_back(trim(field));
|
|
if (comma == std::string_view::npos) {
|
|
break;
|
|
}
|
|
first = comma + 1U;
|
|
}
|
|
return fields;
|
|
}
|
|
|
|
void remove_trailing_empty_fields(std::vector<std::string>& fields) {
|
|
while (!fields.empty() && fields.back().empty()) {
|
|
fields.pop_back();
|
|
}
|
|
}
|
|
|
|
ReferenceCsvReadResult failure(
|
|
const DiagnosticStage stage,
|
|
std::string code,
|
|
std::string message,
|
|
const std::filesystem::path& path,
|
|
const std::size_t line) {
|
|
std::vector<Diagnostic> diagnostics;
|
|
diagnostics.push_back({
|
|
stage,
|
|
Severity::error,
|
|
std::move(code),
|
|
std::move(message),
|
|
SourceLocation{path, line, line == 0U ? 0U : 1U},
|
|
});
|
|
return {{}, std::move(diagnostics)};
|
|
}
|
|
|
|
ReferenceCsvReadResult validation_failure(
|
|
std::string code,
|
|
std::string message,
|
|
const std::filesystem::path& path,
|
|
const std::size_t line) {
|
|
return failure(
|
|
DiagnosticStage::validation,
|
|
std::move(code),
|
|
std::move(message),
|
|
path,
|
|
line);
|
|
}
|
|
|
|
bool is_valid_utf8(const std::string_view text) {
|
|
std::size_t index = 0U;
|
|
while (index < text.size()) {
|
|
const auto first = static_cast<unsigned char>(text[index]);
|
|
if (first <= 0x7FU) {
|
|
++index;
|
|
continue;
|
|
}
|
|
|
|
std::size_t continuation_count = 0U;
|
|
std::uint32_t code_point = 0U;
|
|
std::uint32_t minimum = 0U;
|
|
if (first >= 0xC2U && first <= 0xDFU) {
|
|
continuation_count = 1U;
|
|
code_point = first & 0x1FU;
|
|
minimum = 0x80U;
|
|
} else if (first >= 0xE0U && first <= 0xEFU) {
|
|
continuation_count = 2U;
|
|
code_point = first & 0x0FU;
|
|
minimum = 0x800U;
|
|
} else if (first >= 0xF0U && first <= 0xF4U) {
|
|
continuation_count = 3U;
|
|
code_point = first & 0x07U;
|
|
minimum = 0x10000U;
|
|
} else {
|
|
return false;
|
|
}
|
|
if (index + continuation_count >= text.size()) {
|
|
return false;
|
|
}
|
|
for (std::size_t offset = 1U; offset <= continuation_count; ++offset) {
|
|
const auto continuation =
|
|
static_cast<unsigned char>(text[index + offset]);
|
|
if ((continuation & 0xC0U) != 0x80U) {
|
|
return false;
|
|
}
|
|
code_point = (code_point << 6U) | (continuation & 0x3FU);
|
|
}
|
|
if (code_point < minimum || code_point > 0x10FFFFU ||
|
|
(code_point >= 0xD800U && code_point <= 0xDFFFU)) {
|
|
return false;
|
|
}
|
|
index += continuation_count + 1U;
|
|
}
|
|
return true;
|
|
}
|
|
|
|
std::size_t line_at_offset(
|
|
const std::string_view text,
|
|
const std::size_t offset) {
|
|
return 1U + static_cast<std::size_t>(std::ranges::count(
|
|
text.substr(0U, offset), '\n'));
|
|
}
|
|
|
|
std::vector<std::string_view> required_columns(
|
|
const ReferenceQuantity quantity) {
|
|
switch (quantity) {
|
|
case ReferenceQuantity::displacement:
|
|
return {
|
|
node_column,
|
|
"U-U1",
|
|
"U-U2",
|
|
"U-U3",
|
|
"UR-UR1",
|
|
"UR-UR2",
|
|
"UR-UR3",
|
|
};
|
|
case ReferenceQuantity::reaction:
|
|
return {
|
|
node_column,
|
|
"RF-RF1",
|
|
"RF-RF2",
|
|
"RF-RF3",
|
|
"RM-RM1",
|
|
"RM-RM2",
|
|
"RM-RM3",
|
|
};
|
|
case ReferenceQuantity::internal_force:
|
|
return {
|
|
element_column,
|
|
node_column,
|
|
"SF-SF1",
|
|
"SF-SF2",
|
|
"SF-SF3",
|
|
"SM-SM1",
|
|
"SM-SM2",
|
|
"SM-SM3",
|
|
};
|
|
case ReferenceQuantity::centroid_stress:
|
|
return {element_column, node_column, "Sxx"};
|
|
}
|
|
return {};
|
|
}
|
|
|
|
std::vector<std::string_view> value_columns(
|
|
const ReferenceQuantity quantity) {
|
|
switch (quantity) {
|
|
case ReferenceQuantity::displacement:
|
|
return {"U-U1", "U-U2", "U-U3", "UR-UR1", "UR-UR2", "UR-UR3"};
|
|
case ReferenceQuantity::reaction:
|
|
return {"RF-RF1", "RF-RF2", "RF-RF3", "RM-RM1", "RM-RM2", "RM-RM3"};
|
|
case ReferenceQuantity::internal_force:
|
|
return {"SF-SF1", "SF-SF3", "SF-SF2", "SM-SM3", "SM-SM1", "SM-SM2"};
|
|
case ReferenceQuantity::centroid_stress:
|
|
return {"Sxx"};
|
|
}
|
|
return {};
|
|
}
|
|
|
|
bool parse_positive_label(
|
|
std::string_view text,
|
|
std::int64_t& value) {
|
|
if (text.starts_with('+')) {
|
|
text.remove_prefix(1U);
|
|
}
|
|
const auto parsed = std::from_chars(
|
|
text.data(), text.data() + text.size(), value);
|
|
return !text.empty() && parsed.ec == std::errc{} &&
|
|
parsed.ptr == text.data() + text.size() && value > 0;
|
|
}
|
|
|
|
bool parse_finite_double(std::string_view text, double& value) {
|
|
if (text.starts_with('+')) {
|
|
text.remove_prefix(1U);
|
|
}
|
|
const auto parsed = std::from_chars(
|
|
text.data(),
|
|
text.data() + text.size(),
|
|
value,
|
|
std::chars_format::general);
|
|
return !text.empty() && parsed.ec == std::errc{} &&
|
|
parsed.ptr == text.data() + text.size() &&
|
|
std::isfinite(value);
|
|
}
|
|
|
|
} // namespace
|
|
|
|
ReferenceCsvReadResult read_reference_csv(
|
|
const ReferenceQuantity quantity,
|
|
const std::filesystem::path& path,
|
|
const std::string_view single_instance_name) {
|
|
std::ifstream file{path, std::ios::binary};
|
|
if (!file) {
|
|
return failure(
|
|
DiagnosticStage::io,
|
|
"validation.reference_csv_open_failed",
|
|
"Unable to open reference CSV file.",
|
|
path,
|
|
0U);
|
|
}
|
|
|
|
std::string bytes{
|
|
std::istreambuf_iterator<char>{file},
|
|
std::istreambuf_iterator<char>{},
|
|
};
|
|
if (file.bad()) {
|
|
return failure(
|
|
DiagnosticStage::io,
|
|
"validation.reference_csv_read_failed",
|
|
"Failed while reading reference CSV file.",
|
|
path,
|
|
0U);
|
|
}
|
|
|
|
constexpr std::string_view bom{"\xEF\xBB\xBF"};
|
|
if (bytes.starts_with(bom)) {
|
|
bytes.erase(0U, bom.size());
|
|
}
|
|
const std::size_t misplaced_bom = bytes.find(bom);
|
|
if (misplaced_bom != std::string::npos || !is_valid_utf8(bytes)) {
|
|
return validation_failure(
|
|
"validation.reference_csv_invalid_encoding",
|
|
"Reference CSV must be UTF-8 with an optional BOM only at the file start.",
|
|
path,
|
|
misplaced_bom == std::string::npos
|
|
? 1U
|
|
: line_at_offset(bytes, misplaced_bom));
|
|
}
|
|
|
|
std::istringstream input{bytes};
|
|
std::string line;
|
|
if (!std::getline(input, line)) {
|
|
return validation_failure(
|
|
"validation.reference_csv_missing_header",
|
|
"Reference CSV is missing its header row.",
|
|
path,
|
|
1U);
|
|
}
|
|
|
|
std::vector<std::string> headers = split_fields(line);
|
|
remove_trailing_empty_fields(headers);
|
|
ColumnIndices column_indices;
|
|
for (std::size_t index = 0U; index < headers.size(); ++index) {
|
|
if (headers[index].empty() ||
|
|
!column_indices.emplace(headers[index], index).second) {
|
|
return validation_failure(
|
|
"validation.reference_csv_duplicate_column",
|
|
"Reference CSV contains an empty or duplicate column name.",
|
|
path,
|
|
1U);
|
|
}
|
|
}
|
|
|
|
const std::vector<std::string_view> required = required_columns(quantity);
|
|
for (const std::string_view name : required) {
|
|
if (!column_indices.contains(name)) {
|
|
return validation_failure(
|
|
"validation.reference_csv_missing_column",
|
|
"Reference CSV is missing required column '" +
|
|
std::string{name} + "'.",
|
|
path,
|
|
1U);
|
|
}
|
|
}
|
|
const bool has_instance = column_indices.contains(instance_column);
|
|
if (!has_instance && single_instance_name.empty()) {
|
|
return validation_failure(
|
|
"validation.reference_csv_missing_column",
|
|
"Reference CSV omits 'Part Instance Name' without a single-Instance name.",
|
|
path,
|
|
1U);
|
|
}
|
|
|
|
std::set<std::string, std::less<>> allowed_columns;
|
|
allowed_columns.emplace(instance_column);
|
|
for (const std::string_view name : required) {
|
|
allowed_columns.emplace(name);
|
|
}
|
|
for (const std::string& header : headers) {
|
|
if (!allowed_columns.contains(header)) {
|
|
return validation_failure(
|
|
"validation.reference_csv_unsupported_column",
|
|
"Reference CSV contains unsupported column '" + header + "'.",
|
|
path,
|
|
1U);
|
|
}
|
|
}
|
|
|
|
const std::vector<std::string_view> components = value_columns(quantity);
|
|
std::vector<ReferenceRow> rows;
|
|
std::set<PositionKey> positions;
|
|
std::size_t line_number = 1U;
|
|
while (std::getline(input, line)) {
|
|
++line_number;
|
|
if (trim(line).empty()) {
|
|
continue;
|
|
}
|
|
std::vector<std::string> fields = split_fields(line);
|
|
remove_trailing_empty_fields(fields);
|
|
if (fields.size() != headers.size()) {
|
|
return validation_failure(
|
|
"validation.reference_csv_invalid_row",
|
|
"Reference CSV row field count does not match the header.",
|
|
path,
|
|
line_number);
|
|
}
|
|
|
|
std::string instance_name = has_instance
|
|
? fields[column_index(
|
|
column_indices,
|
|
instance_column)]
|
|
: std::string{single_instance_name};
|
|
if (instance_name.empty()) {
|
|
return validation_failure(
|
|
"validation.reference_csv_invalid_row",
|
|
"Reference CSV row has an empty Instance name.",
|
|
path,
|
|
line_number);
|
|
}
|
|
|
|
std::int64_t entity_label = 0;
|
|
const std::string_view entity_column =
|
|
quantity == ReferenceQuantity::displacement ||
|
|
quantity == ReferenceQuantity::reaction
|
|
? node_column
|
|
: element_column;
|
|
if (!parse_positive_label(
|
|
fields[column_index(column_indices, entity_column)],
|
|
entity_label)) {
|
|
return validation_failure(
|
|
"validation.reference_csv_invalid_number",
|
|
"Reference CSV entity label must be a positive integer.",
|
|
path,
|
|
line_number);
|
|
}
|
|
|
|
std::optional<std::int64_t> end_node_label;
|
|
if (quantity == ReferenceQuantity::internal_force ||
|
|
quantity == ReferenceQuantity::centroid_stress) {
|
|
std::int64_t parsed_end_node = 0;
|
|
if (!parse_positive_label(
|
|
fields[column_index(column_indices, node_column)],
|
|
parsed_end_node)) {
|
|
return validation_failure(
|
|
"validation.reference_csv_invalid_number",
|
|
"Reference CSV end-node label must be a positive integer.",
|
|
path,
|
|
line_number);
|
|
}
|
|
end_node_label = parsed_end_node;
|
|
}
|
|
|
|
std::vector<double> values;
|
|
values.reserve(components.size());
|
|
for (const std::string_view component : components) {
|
|
double value = 0.0;
|
|
if (!parse_finite_double(
|
|
fields[column_index(column_indices, component)],
|
|
value)) {
|
|
return validation_failure(
|
|
"validation.reference_csv_invalid_number",
|
|
"Reference CSV component '" + std::string{component} +
|
|
"' must be a finite number.",
|
|
path,
|
|
line_number);
|
|
}
|
|
values.push_back(value);
|
|
}
|
|
|
|
PositionKey key{instance_name, entity_label, end_node_label};
|
|
if (!positions.insert(key).second) {
|
|
return validation_failure(
|
|
"validation.reference_csv_duplicate_row",
|
|
"Reference CSV contains a duplicate result position.",
|
|
path,
|
|
line_number);
|
|
}
|
|
rows.push_back({
|
|
quantity,
|
|
{std::move(instance_name), entity_label, end_node_label},
|
|
std::move(values),
|
|
});
|
|
}
|
|
if (input.bad()) {
|
|
return failure(
|
|
DiagnosticStage::io,
|
|
"validation.reference_csv_read_failed",
|
|
"Failed while reading reference CSV file.",
|
|
path,
|
|
line_number);
|
|
}
|
|
if (rows.empty()) {
|
|
return validation_failure(
|
|
"validation.reference_csv_missing_rows",
|
|
"Reference CSV contains no result rows.",
|
|
path,
|
|
line_number);
|
|
}
|
|
return {std::move(rows), {}};
|
|
}
|
|
|
|
} // namespace fesa
|