// Copyright 2017-2020 The Verible Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//      http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

#include "verilog/analysis/checkers/line_length_rule.h"

#include <cstddef>
#include <set>
#include <string>

#include "absl/status/status.h"
#include "absl/strings/match.h"
#include "absl/strings/str_cat.h"
#include "absl/strings/string_view.h"
#include "common/analysis/lint_rule_status.h"
#include "common/strings/comment_utils.h"
#include "common/strings/utf8.h"
#include "common/text/config_utils.h"
#include "common/text/constants.h"
#include "common/text/text_structure.h"
#include "common/text/token_info.h"
#include "common/text/token_stream_view.h"
#include "common/util/iterator_range.h"
#include "verilog/analysis/descriptions.h"
#include "verilog/analysis/lint_rule_registry.h"
#include "verilog/analysis/verilog_linter_constants.h"
#include "verilog/parser/verilog_token_classifications.h"
#include "verilog/parser/verilog_token_enum.h"

namespace verilog {
namespace analysis {

using verible::LintRuleStatus;
using verible::LintViolation;
using verible::TextStructureView;
using verible::TokenInfo;
using verible::TokenSequence;

// Register the lint rule
VERILOG_REGISTER_LINT_RULE(LineLengthRule);

static constexpr absl::string_view kMessage = "Line length exceeds max: ";

#if 0  // See comment below about comment-reflowing being implemented
static bool ContainsAnyWhitespace(absl::string_view s) {
  for (char c : s) {
    if (absl::ascii_isspace(c)) return true;
  }
  return false;
}
#endif

const LintRuleDescriptor& LineLengthRule::GetDescriptor() {
  static const LintRuleDescriptor d{
      .name = "line-length",
      .topic = "line-length",
      .desc =
          "Checks that all lines do not exceed the maximum allowed "
          "length. ",
      .param = {{"length", absl::StrCat(kDefaultLineLength),
                 "Desired line length"}},
  };
  return d;
}

// Returns true if line is an exceptional case that should allow excessive
// length.
static bool AllowLongLineException(TokenSequence::const_iterator token_begin,
                                   TokenSequence::const_iterator token_end) {
  // There may be no tokens on this line if the lexer skipped them.
  // TODO(b/134180314): Preserve all text in lexer.
  if (token_begin == token_end) return true;  // Conservatively ignore.
  auto last_token = token_end - 1;            // Point to last token.
  if (last_token > token_begin && last_token->token_enum() == verible::TK_EOF) {
    --last_token;
  }

  // Point to last non-newline.
  if (last_token > token_begin && last_token->token_enum() == TK_NEWLINE) {
    --last_token;
  }

  // Ignore leading whitespace, to find first non-space token.
  while (token_begin < token_end && token_begin->token_enum() == TK_SPACE) {
    ++token_begin;
  }

  // Single token case:
  // If there is only one token on this line, forgive non-comment tokens,
  // but examine comment tokens deeper.
  if (token_begin == last_token) {
    switch (token_begin->token_enum()) {
      case TK_EOL_COMMENT: {
        // TODO(b/72010240): formatter: reflow comments
        // Ideally, a comment whose contents can be split on spaces
        // should be reflowed to spill onto new commented lines.
        // However the formatter hasn't implemented this yet, and comments
        // remain as atomic tokens, so fixing comment indentation may cause
        // line-length violations.  The compromise for now is to forgive
        // this case (no matter what the length).
        return true;

        // Once comment-reflowing is implemented, re-enable the following:
        // If comment consist of more than one token, it should be split.
        // const absl::string_view comment_contents =
        //     verible::StripCommentAndSpacePadding(token_begin->text);
        // return !ContainsAnyWhitespace(comment_contents);
      }
      // TODO(fangism): examine "long string literals"
      // TODO(fangism): case TK_COMMENT_BLOCK:
      // Multi-line comments need deeper inspection.
      default:
        return true;
    }
  }

  // Multi-token cases:
  switch (token_begin->token_enum()) {
    case PP_include:
      // TODO(fangism): Could try to be more specific and inspect this line's
      // tokens further, but it is acceptable to forgive all `include lines.
      return true;
    case PP_ifdef:
    case PP_ifndef:
    case PP_endif:
      // Include guards (if they reflect the full path) can be long.
      // TODO(fangism): Could examine lines further and determine whether or
      // not length could have been reduced, but not bothering for now.
      return true;
    // TODO(fangism): Consider whether or not PP_else and PP_elsif should
    // be exempt from length checks as well.
    default:
      break;
  }

  if (IsComment(verilog_tokentype(last_token->token_enum()))) {
    // Check for end-of-line comment that contain lint waivers.
    const absl::string_view text =
        verible::StripCommentAndSpacePadding(last_token->text());
    if (absl::StartsWith(text, "ri lint_check_waive")) {
      // TODO(fangism): Could make this pattern more space-insensitive
      return true;
    }
    if (absl::StartsWith(text, kLinterTrigger)) {
      // This is the waiver for this linter tool.
      // verible/verilog/tools/lint/README.md
      return true;
    }
    // TODO(fangism): add "noformat" formatter directives.
  }

  return false;
}

void LineLengthRule::Lint(const TextStructureView& text_structure,
                          absl::string_view) {
  size_t lineno = 0;
  for (const auto& line : text_structure.Lines()) {
    const int observed_line_length = verible::utf8_len(line);
    if (observed_line_length > line_length_limit_) {
      const auto token_range = text_structure.TokenRangeOnLine(lineno);
      // Recall that token_range is *unfiltered* and may contain non-essential
      // whitespace 'tokens'.
      if (!AllowLongLineException(token_range.begin(), token_range.end())) {
        // Fake a token that marks the offending range of text.
        TokenInfo token(TK_OTHER, line.substr(line_length_limit_));
        const std::string msg = absl::StrCat(kMessage, line_length_limit_,
                                             "; is: ", observed_line_length);
        violations_.insert(LintViolation(token, msg));
      }
    }
    ++lineno;
  }
}

absl::Status LineLengthRule::Configure(absl::string_view configuration) {
  using verible::config::SetInt;
  return verible::ParseNameValues(
      configuration, {{"length", SetInt(&line_length_limit_, kMinimumLineLength,
                                        kMaximumLineLength)}});
}

LintRuleStatus LineLengthRule::Report() const {
  return LintRuleStatus(violations_, GetDescriptor());
}

}  // namespace analysis
}  // namespace verilog